CN108304425A - A kind of graph text information recommends method, apparatus and system - Google Patents

A kind of graph text information recommends method, apparatus and system Download PDF

Info

Publication number
CN108304425A
CN108304425A CN201710264284.9A CN201710264284A CN108304425A CN 108304425 A CN108304425 A CN 108304425A CN 201710264284 A CN201710264284 A CN 201710264284A CN 108304425 A CN108304425 A CN 108304425A
Authority
CN
China
Prior art keywords
text information
graph text
instalments
graph
candidate
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201710264284.9A
Other languages
Chinese (zh)
Other versions
CN108304425B (en
Inventor
沈宇亮
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen Yayue Technology Co ltd
Original Assignee
Tencent Technology Shenzhen Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tencent Technology Shenzhen Co Ltd filed Critical Tencent Technology Shenzhen Co Ltd
Priority to CN201710264284.9A priority Critical patent/CN108304425B/en
Publication of CN108304425A publication Critical patent/CN108304425A/en
Application granted granted Critical
Publication of CN108304425B publication Critical patent/CN108304425B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/957Browsing optimisation, e.g. caching or content distillation
    • G06F16/9577Optimising the visualization of content, e.g. distillation of HTML documents
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/953Querying, e.g. by the use of web search engines
    • G06F16/9535Search customisation based on user profiles and personalisation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00Handling natural language data
    • G06F40/20Natural language analysis
    • G06F40/279Recognition of textual entities
    • G06F40/289Phrasal analysis, e.g. finite state techniques or chunking

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Information Transfer Between Computers (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The embodiment of the invention discloses a kind of graph text informations to recommend method, apparatus and system;The embodiment of the present invention receives the graph text information that terminal is sent and obtains request, the graph text information obtains request and carries graph text information mark, then, it is obtained and is asked according to the graph text information, the graph text information, which is sent, to the terminal identifies corresponding target graph text information, determine whether the target graph text information is to publish in instalments graph text information, if, this is then obtained from candidate graph text information set publish in instalments the follow-up of graph text information publish in instalments graph text information, the follow-up Periodical front cover information for publishing in instalments graph text information is sent to the terminal, to recommend this subsequently to publish in instalments graph text information;The program can promote the transmission speed of graph text information and the displaying speed of graph text information, and then improve the promptness of graph text information displaying and the reading experience of user.

Description

A kind of graph text information recommends method, apparatus and system
Technical field
The present invention relates to communication technique fields, and in particular to a kind of graph text information recommends method, apparatus and system.
Background technology
With the development of Internet technology, network is increasingly becoming the source that people obtain information.Therefore, it reads currently on the market Class is read to apply for popular with users.Article information can be obtained from network side by reading class application, and show the text obtained to user in fact Chapter information.
Specifically, the server for reading class application can be according to the public platform and user interest that user subscribes to, to corresponding Article is pushed to corresponding terminal, and the terminal is by reading class application displaying this article, so that user reads.
It is picture and text form that current article is most of comprising picture and the corresponding word of picture.Therefore, general clothes The article information substantially graph text information of business device push.However, some graph text informations contain many in magnanimity graph text information Picture, the data volume of these graph text informations is bigger, the transmission of server information when sending these graph text informations to terminal Speed can be slow, reduces the promptness of graph text information displaying, while a large amount of picture can also cause bad read to user Read experience.
Invention content
A kind of graph text information of offer of the embodiment of the present invention recommends method, apparatus and system, can improve the hair of graph text information The promptness for sending speed and graph text information to show.
The embodiment of the present invention provides a kind of graph text information recommendation method, including:
It receives the graph text information that terminal is sent and obtains request, the graph text information obtains request and carries graph text information mark;
It is obtained and is asked according to the graph text information, sending the graph text information to the terminal identifies corresponding target picture and text Information;
Determine whether the target graph text information is to publish in instalments graph text information;
If so, publishing in instalments the follow-up of graph text information described in being obtained from candidate graph text information set publishes in instalments graph text information;
The follow-up Periodical front cover information for publishing in instalments graph text information is sent to the terminal, to recommend the follow-up picture and text of publishing in instalments to believe Breath.
Correspondingly, the embodiment of the present invention also provides a kind of graph text information recommendation apparatus, including:
Receiving unit, the graph text information for receiving terminal transmission obtain request, and the graph text information obtains request and carries Graph text information identifies;
Graph text information transmission unit is asked for being obtained according to the graph text information, and the picture and text are sent to the terminal The corresponding target graph text information of message identification;
Determination unit, for determining whether the target graph text information is to publish in instalments graph text information;
Publish in instalments acquiring unit, for the determination unit determine the target graph text information be publish in instalments graph text information when, from The follow-up of graph text information is published in instalments described in being obtained in candidate graph text information set publishes in instalments graph text information;
Periodical front cover information transmission unit, for sending the follow-up Periodical front cover information for publishing in instalments graph text information to the terminal, with Recommend described subsequently to publish in instalments graph text information.
Correspondingly, the embodiment of the present invention also provides a kind of graph text information commending system, including it is provided in an embodiment of the present invention Any graph text information recommendation apparatus.
The embodiment of the present invention receives the graph text information that terminal is sent and obtains request, which obtains request and carry picture and text Then message identification is obtained according to the graph text information and is asked, sending the graph text information to the terminal identifies corresponding target picture and text Information determines whether the target graph text information is to publish in instalments graph text information, if so, obtaining the company from candidate graph text information set It carries the follow-up of graph text information and publishes in instalments graph text information, the follow-up Periodical front cover information for publishing in instalments graph text information is sent to the terminal, to recommend This subsequently publishes in instalments graph text information.The program, which can identify, publishes in instalments graph text information, and publishes in instalments picture and text by the follow-up of graph text information is published in instalments Information recommendation comes out;Using the program may be implemented by comprising a large amount of pictures graph text information split into several publish in instalments picture and text letter This after sending some to terminal and publishing in instalments graph text information is published in instalments the follow-up of graph text information and publishes in instalments graph text information and recommend by breath Come, user is facilitated to read complete original graph text information;Therefore, the program can promote the transmission speed and figure of graph text information The displaying speed of literary information, and then improve the promptness of graph text information displaying and the reading experience of user.
Description of the drawings
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for For those skilled in the art, without creative efforts, it can also be obtained according to these attached drawings other attached Figure.
Fig. 1 a are the schematic diagram of a scenario of graph text information commending system provided in an embodiment of the present invention;
Fig. 1 b are the flow diagrams that graph text information provided in an embodiment of the present invention recommends method;
Fig. 2 a are another flow diagrams that graph text information provided in an embodiment of the present invention recommends method;
Fig. 2 b are the schematic diagrames at graph text information displaying provided in an embodiment of the present invention interface;
Fig. 2 c are another schematic diagrames at graph text information displaying provided in an embodiment of the present invention interface;
Fig. 2 d are recommendation graph text information displaying schematic diagrames provided in an embodiment of the present invention;
Fig. 3 a are the first structural schematic diagrams of graph text information recommendation apparatus provided in an embodiment of the present invention;
Fig. 3 b are second of structural schematic diagrams of graph text information recommendation apparatus provided in an embodiment of the present invention;
Fig. 3 c are the third structural schematic diagrams of graph text information recommendation apparatus provided in an embodiment of the present invention;
Fig. 3 d are the 4th kind of structural schematic diagrams of graph text information recommendation apparatus provided in an embodiment of the present invention;
Fig. 4 is the structural schematic diagram of server provided in an embodiment of the present invention.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on Embodiment in the present invention, the every other implementation that those skilled in the art are obtained without creative efforts Example, shall fall within the protection scope of the present invention.
An embodiment of the present invention provides a kind of graph text informations to recommend method, apparatus and system.
An embodiment of the present invention provides a kind of graph text information commending systems, and referring to Fig. 1 a, which includes that the present invention is implemented Any graph text information recommendation apparatus that example is provided, the graph text information recommendation apparatus can be integrated in the network equipments such as server In.In addition, graph text information commending system can also include other equipment, such as user equipment (such as terminal).
Wherein, user equipment, which is used to send graph text information to service, obtains request, and the graph text information obtains request and carries Graph text information identifies;And show target graph text information and Periodical front cover information, etc. that server returns.
Server is used to obtain request according to the graph text information and be corresponded to the terminal transmission graph text information mark Target graph text information, determine whether the target graph text information is to publish in instalments graph text information;If so, from candidate graph text information collection The follow-up of graph text information is published in instalments described in being obtained in conjunction publishes in instalments graph text information;Subsequently picture and text are published in instalments to user equipment transmission is described The Periodical front cover information of information, to recommend described subsequently to publish in instalments graph text information.
It will be described in detail respectively below.
Embodiment one,
The present embodiment will be described from the angle of graph text information recommendation apparatus, which specifically can be with It is integrated in the equipment such as server.
A kind of graph text information recommendation method, including:It receives the graph text information that terminal is sent and obtains request, which obtains It takes request to carry graph text information mark, then, is obtained and asked according to the graph text information, graph text information mark is sent to the terminal Corresponding target graph text information determines whether the target graph text information is to publish in instalments graph text information, if so, from candidate graph text information This is obtained in set publishing in instalments the follow-up of graph text information and publish in instalments graph text information, the follow-up envelope for publishing in instalments graph text information is sent to the terminal Face information, to recommend this subsequently to publish in instalments graph text information.
As shown in Figure 1 b, which recommends the detailed process of method can be as follows:
101, it receives the graph text information that terminal is sent and obtains request, which obtains request and carry graph text information mark Know.
Wherein, graph text information is identified as the unique mark of graph text information, and graph text information mark can be made of character.Than Such as, graph text information mark can be the URL (Uniform Resource Locator, uniform resource locator) of graph text information Deng.
In the present embodiment, graph text information includes pictorial information and text information;For example, graph text information can be atlas, The atlas can be the content-form that a kind of user reads, which may include the explanatory note of picture and picture.
In practical application, terminal can send graph text information to server by application program and obtain request, at this point, service Device receives the request.
102, it is obtained and is asked according to the graph text information, sending the graph text information to the terminal identifies corresponding target picture and text letter Breath.
Specifically, it after receiving graph text information and obtaining request, can be identified from data source (such as this according to graph text information Ground database or other network equipments) it obtains the graph text information and identifies corresponding target graph text information and be then sent to end End, terminal can show that the target graph text information, such user can browse the target figure after receiving target graph text information Literary information.
103, determine whether the target graph text information is to publish in instalments graph text information, if so, 104 are thened follow the steps, if not, holding Row step 106.
For example, can determine whether the target graph text information is company during sending target graph text information to terminal Carry graph text information.
Wherein, the graph text information that graph text information is the form of publishing in instalments is published in instalments, is referred to:Some graph text information splits into multiple sons Sub- graph text information when graph text information is issued several times.For example, when some graph text information entitled " XXXXX (on) ", " XXXXX (under) ", " XXXXX (upper collection) ", " XXXXX (next part) ", " XXXXX (the first collection) " or " XXXXX (the second collection) " When, which is to publish in instalments graph text information.
Optionally, it can generally be found out from title due to publishing in instalments graph text information, the present embodiment can pass through figure The title of literary information determines whether to publish in instalments graph text information.Specifically, step " determines whether the target graph text information is to publish in instalments Graph text information " may include:
Obtain the title word segmentation result set of the target graph text information;
Judge in the title word segmentation result set to whether there is and the default word for publishing in instalments Keywords matching;
If so, determining that the target graph text information is to publish in instalments graph text information.
Wherein, title word segmentation result collection is combined into carries out the participle set obtained after word segmentation processing to the title of graph text information. In the present embodiment, before receiving graph text information and obtaining request, word segmentation processing is carried out to the title of graph text information in advance, then It is saved in corresponding database.At this point it is possible to segment knot according to graph text information mark to obtain the title of target graph text information Fruit set;I.e. step " the title word segmentation result set for obtaining the target graph text information " may include:It is identified according to graph text information The title word segmentation result set of target graph text information is extracted from associated databases.
It in some other implementations, can also be after receiving graph text information and obtaining request, to graph text information Title carries out word segmentation processing, to obtain word segmentation result set;At this point, step " obtains the title participle knot of the target graph text information Fruit set " may include:Word segmentation processing is carried out to the title of graph text information, to obtain word segmentation result set.
Preferably, the present embodiment may be used such as jieba (stammerer) participle mode and be segmented to title.
Wherein, default keyword of publishing in instalments can publish in instalments pass to be preconfigured before receiving graph text information and obtaining request Keyword, this, which publishes in instalments keyword, to publish in instalments word for general use, such as "upper", "lower", " upper collection ", " next part ", " first collects " ", " " the second collection " etc., it can specifically be set according to actual demand.
In the present embodiment, when existing in the title word segmentation result set of graph text information and the default word for publishing in instalments Keywords matching When, then it is assumed that the graph text information is to publish in instalments graph text information, otherwise it is assumed that the graph text information publishes in instalments graph text information to be non-.
For example, graph text information it is entitled " face be worth quick-fried table they (on) " when, after being segmented to title, can be marked Inscribe word segmentation result set { " face value ", " quick-fried table ", " they ", "upper" };Assuming that publish in instalments keyword be "upper", "lower" when, at this point, should Exist in title word segmentation result set and publish in instalments the word i.e. "upper" of Keywords matching, then will think that the graph text information is to publish in instalments Graph text information.
104, this is obtained from candidate graph text information set publish in instalments the follow-up of graph text information publish in instalments graph text information.
Wherein, the follow-up graph text information of publishing in instalments for publishing in instalments graph text information is to believe with publishing in instalments graph text information and belong to the same picture and text Breath, and there is the graph text information of ordinal relation with graph text information is published in instalments.The ordinal relation includes that front and back, upper and lower, left and right etc. are suitable Order relation.For example, when there are the graph text information b of the graph text information a of entitled " XXXXX (on) ", entitled " XXXXX (under) " When, graph text information b is that graph text information is published in instalments in the follow-up of image information a.
For another example, when there are the graph text information c of entitled " XXXXX (first collection) ", entitled " XXXXX (the second collection) " When graph text information d, graph text information d is that graph text information is published in instalments in the follow-up of image information c.
Preferably, it is contemplated that the author for publishing in instalments graph text information and its follow-up continuous picture and text information has relevance, such as may For the same author, therefore, the present embodiment can be tied by the author information of graph text information come preliminary screening, then from screening It is further chosen in fruit and obtains publishing in instalments the follow-up of graph text information and publish in instalments graph text information.Specifically, step is " from candidate graph text information collection This obtained in conjunction publish in instalments the follow-up of graph text information publish in instalments graph text information " may include:
Obtain in candidate's graph text information set candidate graph text information author information and this publish in instalments the work of graph text information Person's information;
According to the author information of candidate's graph text information and this publish in instalments the author information of graph text information, from candidate's picture and text Information aggregate chooses corresponding candidate graph text information, obtains author's correlation candidate graph text information set;
This is chosen from author's correlation candidate graph text information set publish in instalments the follow-up of graph text information publish in instalments graph text information.
For example, according to the author information of candidate's graph text information and this publish in instalments the author information of graph text information, from the time The candidate graph text information for selecting graph text information set to choose and publish in instalments the same author of graph text information;Alternatively, from candidate's graph text information Set, which chooses author and publishes in instalments graph text information author, has incidence relation (author of such as same media, the author of same department Etc.) candidate graph text information.
Publishes in instalments keyword in view of subsequently publishing in instalments the title of graph text information and can generally carry, and it is follow-up publish in instalments graph text information with The graph text information type published in instalments belonging to graph text information is related, the present embodiment can based on publish in instalments keyword and graph text information type come It chooses candidate and publishes in instalments graph text information, is i.e. step " chooses this from author's correlation candidate graph text information set and publishes in instalments graph text information Follow-up publish in instalments graph text information " may include:
Title in author's correlation candidate graph text information set the default candidate graph text information for publishing in instalments keyword is not included into Removal, candidate graph text information set after being removed;
According to the graph text information type of candidate graph text information in candidate graph text information set after removal and this publish in instalments picture and text The graph text information type of information chooses corresponding candidate graph text information from candidate graph text information set after removal, obtains type Correlation candidate graph text information set;
This is chosen from the type correlation candidate graph text information set publish in instalments the follow-up of graph text information publish in instalments graph text information.
For example, the graph text information type for publishing in instalments graph text information with this can be chosen from candidate graph text information set after removal Identical candidate's graph text information, i.e., the same type of candidate of graph text information is found out and published in instalments to candidate graph text information set after removal Graph text information.
Preferably, after obtaining type correlation candidate graph text information set, each candidate picture and text letter in set can be based on The label (tag) of breath subsequently publishes in instalments graph text information to choose.Label refers to:The key of graph text information main contents can be summarized Word.One graph text information can there are one or multiple labels.
Specifically, step " chooses this from the type correlation candidate graph text information set and publishes in instalments subsequently connecting for graph text information Carry graph text information " may include:
It obtains in type correlation candidate graph text information set the label of candidate graph text information and this publishes in instalments graph text information Label;
According to the label of candidate's graph text information and this publish in instalments the label of graph text information, believe from type correlation candidate picture and text This is chosen in breath set publish in instalments the follow-up of graph text information publish in instalments graph text information.
For example, can be chosen from type correlation candidate graph text information set quantity is overlapped with the tag for publishing in instalments graph text information Most candidate graph text information, using as subsequently publishing in instalments graph text information.Specifically, the label and class of graph text information will can be published in instalments The label of candidate graph text information is compared in type correlation candidate graph text information set, obtains the label of each candidate graph text information Quantity (i.e. same label quantity) is overlapped, then, chooses and overlaps the most candidate picture and text letter of quantity with the tag for publishing in instalments graph text information Breath, using as subsequently publishing in instalments graph text information.
Optionally, graph text information is subsequently published in instalments in order to find, the present embodiment can be obtained receiving graph text information Before request, pretreatments of some series are carried out, such as is classified to graph text information and extracts the label of graph text information. Namely before receiving the graph text information that terminal is sent and obtaining request, the graph text information of the present embodiment recommends method that can also wrap It includes:
Tag extraction is carried out to the graph text information in graph text information set, obtains each graph text information in graph text information set Label;
Classify to the graph text information in the graph text information set, obtains each graph text information in graph text information set Graph text information type.
Wherein, which can be selected according to actual demand, for example, the graph text information set may include clothes All graph text informations on business device, the graph text information set are greater than or equal to candidate graph text information set.
For example, can to the graph text information in graph text information set carry out LDA (Latent Dirichlet Allocation, Document subject matter generates model) classification, obtain the graph text information type of each graph text information in graph text information set.
Wherein, the extraction of tag is there are many mode, for example, step " to the graph text information in graph text information set into row label Extract " may include:
Obtain the text word segmentation result set and title word segmentation result set of graph text information in graph text information set;
According to the similarity between word in word in title word segmentation result set and the text word segmentation result set, from the mark Label of the corresponding word as the graph text information is chosen in topic word segmentation result set.
The text word segmentation result collection is combined into carries out the participle set obtained after word segmentation processing to the body matter of graph text information. The title word segmentation result information is that the title of graph text information carries out the participle set obtained after word segmentation processing.The present embodiment can be adopted Mode is segmented with jieba, removes word and stop-word.
The present embodiment, the similarity between word and word may include the vector similarity between term vector, as cosine is similar Degree etc..Therefore, step " according to the similarity between word in word in title word segmentation result set and the text word segmentation result set, Label of the corresponding word as the graph text information is chosen from the title word segmentation result set " may include:
Obtain in the text word segmentation result set title in the positive corresponding term vector of cliction and title word segmentation result set The corresponding term vector of word;
According to the positive corresponding term vector of cliction in the text word segmentation result set, obtain the representative word of graph text information text to Amount;
According to the corresponding term vector of heading in title word segmentation result set, similar to the vector between representing term vector Degree, label of the corresponding word as the graph text information is chosen from the title word segmentation result set.
Wherein, there are many modes for obtaining term vector, and word2vector vector training is carried out to word for example, may be used, To obtain the term vector of word.
Specifically, the vectorial phase of the term vector and representation vector of each heading in title word segmentation result set can be calculated Like degree, then, meet the heading of preset condition as tag from vector similarity is chosen.
For example, when vector similarity is cosine similarity, the word of title word segmentation result set can be traversed, is calculated each The term vector of word and the cosine similarity value for representing term vector, when the cosine similarity value is more than predetermined threshold value, it is determined that should Word is the tag of graph text information.Wherein, predetermined threshold value can be set according to actual demand, and such as 0.5.In the present embodiment, two words to The cosine similarity value of amount is bigger closer to 1, shows that the angle of two term vectors is smaller, and then show that two term vectors are more similar And the word of two term vectors is more similar.
In order to accelerate the extraction of tag and improve the accuracy of tag, the present embodiment can be selected from text word segmentation result set A certain number of positive clictions are taken, then, the representation vector of literary message text are obtained according to the positive selected works word of selection;Namely step " obtaining the corresponding term vector of positive cliction in the text word segmentation result set " may include:It is selected from the text word segmentation result set The target text word of preset quantity is taken, and obtains the corresponding term vector of target text word;
Step " according to the positive corresponding term vector of cliction in the text word segmentation result set, obtains the generation of graph text information text Table term vector " may include:The representative term vector of graph text information text is obtained according to the corresponding term vector of target text word.
It, can be according to TF-IDF (the term frequency-of word in text word segmentation result set in the present embodiment Inverse document frequency, word frequency-inverse file frequency)) weight chooses the target text word of preset quantity, than Such as, it is target text word to take the positive cliction of highest 10 of TF-IDF weight.
After obtaining the corresponding term vector of target text word, the corresponding term vector of target text word can be added, and Make normalized, to obtain representing term vector.
105, the follow-up Periodical front cover information for publishing in instalments graph text information is sent to the terminal, to recommend this subsequently to publish in instalments picture and text letter Breath.
Wherein, Periodical front cover information may include:Subsequently publish in instalments cover picture, title and link of graph text information etc..Terminal The cover of graph text information can be subsequently published in instalments according to Periodical front cover information displaying after receiving Periodical front cover information.The link when the user clicks is It can obtain and subsequently publish in instalments graph text information.
106, conventional recommended flowsheet is executed.
For example, recommending to belong to same type or graph text information with similar label to user with graph text information.
From the foregoing, it will be observed that the embodiment of the present invention, which receives the graph text information that terminal is sent, obtains request, graph text information acquisition is asked It asks and carries graph text information mark, then, obtained and asked according to the graph text information, sending graph text information mark to the terminal corresponds to Target graph text information, determine whether the target graph text information is to publish in instalments graph text information, if so, from candidate graph text information set It is middle obtain this and publish in instalments the follow-up of graph text information publish in instalments graph text information, send the follow-up cover letter for publishing in instalments graph text information to the terminal Breath, to recommend this subsequently to publish in instalments graph text information.The program, which can identify, publishes in instalments graph text information, and will publish in instalments the follow-up of graph text information Graph text information is published in instalments to recommend out;It may be implemented the graph text information comprising a large amount of pictures splitting into several companies using the program Graph text information is carried, this is published in instalments into the follow-up of graph text information after sending some to terminal and publishing in instalments graph text information and publishes in instalments graph text information Recommendation comes out, and user is facilitated to read complete original graph text information;Therefore, the program can promote the transmission speed of graph text information And the displaying speed of graph text information, and then improve the promptness of graph text information displaying and the reading experience of user.
Embodiment two,
According to method described in embodiment one, citing is described in further detail below.
The embodiment of the present invention combines system shown in Fig. 1 a so that graph text information recommendation apparatus is integrated in the server as an example Recommend method to introduce graph text information provided by the invention.
As shown in Figure 2 a, a kind of graph text information recommends method, and detailed process can be as follows:
201, terminal to server sends graph text information and obtains request, which obtains request and carry graph text information mark Know.
With reference to figure 2b, Periodical front cover information (cover picture, title, the chain of multiple graph text informations are shown in instant communication applications It connects);The cover link that the graph text information can be clicked when user needs to browse some graph text information, with triggering terminal to service Device sends graph text information and obtains request, which obtains the graph text information mark that request carries the graph text information currently clicked Know.
For example, entitled when the user clicks, " face is worth they of quick-fried table, wilfully falls in love with vollyball!It should very be thumbed up for them!” Cover when, terminal can send the acquisition request of the graph text information to server, which carries that entitled " face is worth she of quick-fried table , wilfully fall in love with vollyball!It should very be thumbed up for them!" graph text information mark.
202, server obtains the corresponding target graph text information of acquisition request graph-text identification according to graph text information, and to terminal Return to the target graph text information.
For example, server can be searched, entitled " face is worth they of quick-fried table, wilfully falls in love with vollyball!It should very be thumbed up for them! (on) " graph-text content the graph-text content is then sent to terminal.Terminal is shown after receiving the graph-text content, reference Fig. 2 c.
203, server obtains the title word segmentation result set of target graph text information.
Wherein, title word segmentation result collection is combined into carries out the participle set obtained after word segmentation processing to the title of graph text information. For example, can be using the participle set obtained after jieba word segmentation processings.
For example, server in advance can carry out the title of graph text information before receiving graph text information and obtaining request Then word segmentation processing is saved in corresponding database, server can be identified according to graph text information and be carried from database at this time Take title word segmentation result set.
For another example, server can at this time segment title, obtain word segmentation result set.
204, server judges to whether there is in the title word segmentation result set and presets the word for publishing in instalments Keywords matching, if It is to then follow the steps 205, if it is not, thening follow the steps 211.
Wherein, default keyword of publishing in instalments can publish in instalments pass to be preconfigured before receiving graph text information and obtaining request Keyword, this, which publishes in instalments keyword, to publish in instalments word for general use, such as "upper", "lower", " upper collection ", " next part ", " first collects " ", " " the second collection " etc., it can specifically be set according to actual demand.
In the present embodiment, when existing in the title word segmentation result set of graph text information and the default word for publishing in instalments Keywords matching When, then it is assumed that the graph text information is to publish in instalments graph text information, otherwise it is assumed that the graph text information publishes in instalments graph text information to be non-.
For example, " face is worth they of quick-fried table, wilfully falls in love with vollyball with entitled!It should very be thumbed up for them!(on) " figure For literary content, after being segmented to title, title word segmentation result set { " face value ", " quick-fried table ", " they ", " love can be obtained On ", " vollyball ", " thumbing up ", "upper";Assuming that publish in instalments keyword be "upper", "lower" when, at this point, in the title word segmentation result set In the presence of with publish in instalments the word i.e. "upper" of Keywords matching, then will think the graph text information be publish in instalments graph text information.
205, server determines that the target graph text information is to publish in instalments graph text information, is chosen from candidate graph text information set M The identical candidate graph text information of graph text information author is published in instalments with this, obtains author's correlation candidate graph text information set M1.
Specifically, the author information of candidate graph text information in candidate graph text information set M can be obtained and publish in instalments picture and text Then the author information of information carries out the author information for publishing in instalments graph text information with the author information of candidate graph text information respectively Compare, obtains the candidate graph text information of same authors.
206, server, which does not include title in author's correlation candidate graph text information set M1 to preset, publishes in instalments keyword Candidate graph text information removal, candidate graph text information set M2 after being removed.
207, server chooses that publish in instalments graph text information type identical with this in the candidate graph text information set M2 to after removal Candidate graph text information, obtain type correlation candidate graph text information set M3.
Optionally, before step 201, server can in advance classify to graph text information, for example, can be to picture and text Information carries out LDA (Latent Dirichlet Allocation, document subject matter generate model) classification etc..In addition, server can To obtain the graph text information type of candidate graph text information in M2, and the graph text information type of graph text information is published in instalments, then, carried out Type compares, and obtains candidate graph text information identical with graph text information type is published in instalments.
208, server chooses and publishes in instalments the tag coincidence numbers of graph text information from type correlation candidate graph text information set M3 The most candidate graph text information of amount publishes in instalments graph text information to publish in instalments the follow-up of graph text information as this.
Optionally, before step 201, server can also carry out tag extractions, the extraction of tag to graph text information in advance Mode can be as follows:
Obtain the text word segmentation result set wordTxt and title word segmentation result set wordTitle of graph text information;
Highest 10 words of TF-IDF weight are chosen from text word segmentation result set wordTxt, obtain wordTxt ';
Word2vector vector training is carried out to heading in wordTitle and the positive clictions of wordTxt ', to To the term vector of heading, the term vector of positive cliction;
It by the term vector addition of vectors of positive cliction in wordTxt ', and normalizes, obtains the representative word of graph text information text Vector;
The heading in wordTitle is traversed, the term vector of each heading is represented into term vector complementation string with this, if Cosine value>0.5, then it is assumed that the heading is the tag of graph text information.
In the tag of extraction graph text information in advance, at this point, server can obtain candidate graph text information in M3 Then the tag and tag for publishing in instalments graph text information carries out the tag of candidate graph text information in the tag and M3 that publish in instalments graph text information Compare, obtains overlapping the most candidate graph text information of quantity with the tag for publishing in instalments graph text information.
209, server sends the follow-up Periodical front cover information for publishing in instalments graph text information to terminal.
Wherein, Periodical front cover information may include:Subsequently publish in instalments cover picture, title and link of graph text information etc..
210, terminal subsequently publishes in instalments the cover of picture and text according to Periodical front cover information displaying.
For example, server is finally found, entitled " face is worth they of quick-fried table, wilfully falls in love with vollyball!Should be very them It thumbs up!(under) " follow-up publish in instalments graph text information;At this point, the Periodical front cover information that this is subsequently published in instalments graph text information by service is sent to end End.With reference to figure 2d, in terminal instant messaging application the follow-up cover for publishing in instalments graph text information can be formed on interface.
211, server executes conventional graph text information recommended flowsheet.
From the foregoing, it will be observed that the embodiment of the present invention, which receives the graph text information that terminal is sent, obtains request, graph text information acquisition is asked It asks and carries graph text information mark, then, obtained and asked according to the graph text information, sending graph text information mark to the terminal corresponds to Target graph text information, determine whether the target graph text information is to publish in instalments graph text information, if so, from candidate graph text information set It is middle obtain this and publish in instalments the follow-up of graph text information publish in instalments graph text information, send the follow-up cover letter for publishing in instalments graph text information to the terminal Breath, to recommend this subsequently to publish in instalments graph text information.The program, which can identify, publishes in instalments graph text information, and will publish in instalments the follow-up of graph text information Graph text information is published in instalments to recommend out;It may be implemented the graph text information comprising a large amount of pictures splitting into several companies using the program Graph text information is carried, this is published in instalments into the follow-up of graph text information after sending some to terminal and publishing in instalments graph text information and publishes in instalments graph text information Recommendation comes out, and user is facilitated to read complete original graph text information;Therefore, the program can promote the transmission speed of graph text information And the displaying speed of graph text information, and then improve the promptness of graph text information displaying, the conversion ratio of graph text information and user Reading experience.
Embodiment three,
In order to preferably implement above method, the embodiment of the present invention also provides graph text information recommendation apparatus, as shown in Figure 3a, The graph text information recommendation apparatus includes:Receiving unit 301, determination unit 303, publishes in instalments acquisition list at graph text information transmission unit 302 Member 304 and knot Periodical front cover information transmission unit 305, it is as follows:
(1) receiving unit 301;
Receiving unit 301, the graph text information for receiving terminal transmission obtain request, which obtains request and carry Graph text information identifies.
Wherein, graph text information is identified as the unique mark of graph text information, and graph text information mark can be made of character.Than Such as, graph text information mark can be the URL (Uniform Resource Locator, uniform resource locator) of graph text information Deng.
In the present embodiment, graph text information includes pictorial information and text information;For example, graph text information can be atlas, The atlas can be the content-form that a kind of user reads, which may include the explanatory note of picture and picture.
(2) graph text information transmission unit 302;
Graph text information transmission unit 302 asks for being obtained according to the graph text information, the graph text information is sent to the terminal Identify corresponding target graph text information.
For example, graph text information transmission unit 302, can be used for according to graph text information mark from data source (such as local data base, Or other network equipments) obtain the graph text information and identify corresponding target graph text information and be then sent to terminal.
(3) determination unit 303;
Determination unit 303, for determining whether the target graph text information is to publish in instalments graph text information.
Specifically, it is determined that unit 303, can be used for sending target graph text information to terminal with unit in graph text information transmission During, determine whether the target graph text information is to publish in instalments graph text information.
Wherein, the graph text information that graph text information is the form of publishing in instalments is published in instalments, is referred to:Some graph text information splits into multiple sons Sub- graph text information when graph text information is issued several times.For example, when some graph text information entitled " XXXXX (on) ", " XXXXX (under) ", " XXXXX (upper collection) ", " XXXXX (next part) ", " XXXXX (the first collection) " or " XXXXX (the second collection) " When, which is to publish in instalments graph text information.
Optionally, it can generally be found out from title due to publishing in instalments graph text information, the present embodiment can pass through figure The title of literary information determines whether to publish in instalments graph text information.For example, with reference to figure 3b, which may include:
Participle obtains subelement 3031, the title word segmentation result set for obtaining the target graph text information;
Judgment sub-unit 3032 publishes in instalments keyword in the title word segmentation result set for judging with the presence or absence of with default The word matched;
Determination subelement 3033, for when the judgment sub-unit 3032 is judged as YES, determining that the target graph text information is Publish in instalments graph text information.
Wherein, title word segmentation result collection is combined into carries out the participle set obtained after word segmentation processing to the title of graph text information.
Wherein, default keyword of publishing in instalments can publish in instalments pass to be preconfigured before receiving graph text information and obtaining request Keyword, this, which publishes in instalments keyword, to publish in instalments word for general use, such as "upper", "lower", " upper collection ", " next part ", " first collects " ", " " the second collection " etc., it can specifically be set according to actual demand.
(4) acquiring unit 304 is published in instalments;
Acquiring unit 304 is published in instalments, for determining that the target graph text information is to publish in instalments graph text information in the determination unit 303 When, this is obtained from candidate graph text information set publish in instalments the follow-up of graph text information publish in instalments graph text information.
Wherein, the follow-up graph text information of publishing in instalments for publishing in instalments graph text information is to believe with publishing in instalments graph text information and belong to the same picture and text Breath, and there is the graph text information of ordinal relation with graph text information is published in instalments.The ordinal relation includes that front and back, upper and lower, left and right etc. are suitable Order relation.For example, when there are the graph text information b of the graph text information a of entitled " XXXXX (on) ", entitled " XXXXX (under) " When, graph text information b is that graph text information is published in instalments in the follow-up of image information a.
Preferably, it is contemplated that the author for publishing in instalments graph text information and its follow-up continuous picture and text information has relevance, such as may For the same author, therefore, the present embodiment can be tied by the author information of graph text information come preliminary screening, then from screening It is further chosen in fruit and obtains publishing in instalments the follow-up of graph text information and publish in instalments graph text information.For example, with reference to figure 3c, acquiring unit is published in instalments 304 may include:
Author information obtains subelement 3041, obtains author's letter of candidate graph text information in candidate's graph text information set Breath and this publish in instalments the author information of graph text information;
Author chooses son single 3042, for according to the author information of candidate's graph text information and this publish in instalments graph text information Author information chooses corresponding candidate graph text information from candidate's graph text information set, obtains author's correlation candidate graph text information Set;
Acquisition of information subelement 3043 is published in instalments, publishes in instalments figure for choosing this from author's correlation candidate graph text information set Publish in instalments graph text information in the follow-up of literary information.
Wherein, acquisition of information subelement 3043 is published in instalments, can be specifically used for:By author's correlation candidate graph text information set Middle title does not include the default candidate graph text information removal for publishing in instalments keyword, candidate graph text information set after being removed;
According to the graph text information type of candidate graph text information in candidate graph text information set after removal and this publish in instalments picture and text The graph text information type of information chooses corresponding candidate graph text information from candidate graph text information set after removal, obtains type Correlation candidate graph text information set;
This is chosen from the type correlation candidate graph text information set publish in instalments the follow-up of graph text information publish in instalments graph text information.
For example, the graph text information type for publishing in instalments graph text information with this can be chosen from candidate graph text information set after removal Identical candidate's graph text information.
Preferably, this publishes in instalments acquisition of information subelement 3043, can be specifically used for obtaining type correlation candidate graph text information In set the label of candidate graph text information and this publish in instalments the label of graph text information;According to the label of candidate's graph text information with And this publishes in instalments the label of graph text information, this is chosen from type correlation candidate graph text information set and publishes in instalments subsequently connecting for graph text information Carry graph text information.
For example, can be chosen from type correlation candidate graph text information set quantity is overlapped with the tag for publishing in instalments graph text information Most candidate graph text information, using as subsequently publishing in instalments graph text information.
(5) Periodical front cover information transmission unit 305;
Periodical front cover information transmission unit 305, for sending the follow-up Periodical front cover information for publishing in instalments graph text information to the terminal, to push away It recommends this and subsequently publishes in instalments graph text information.
Wherein, Periodical front cover information may include:Subsequently publish in instalments cover picture, title and link of graph text information etc..
Optionally, with reference to figure 3d, the present embodiment graph text information recommendation apparatus can also include:
Tag extraction unit 306 is used for before receiving unit 301 receives graph text information acquisition request, to graph text information Graph text information in set carries out tag extraction, obtains the label of each graph text information in graph text information set;
Taxon 307 is used for before receiving unit 301 receives graph text information acquisition request, to the graph text information collection Graph text information in conjunction is classified, and the graph text information type of each graph text information in graph text information set is obtained.
Classify to graph text information for example, IDA models may be used in taxon 307.
For example, tag extraction unit 306 can be used for:
Obtain the text word segmentation result set and title word segmentation result set of graph text information in graph text information set;
According to the similarity between word in word in title word segmentation result set and the text word segmentation result set, from the mark Label of the corresponding word as the graph text information is chosen in topic word segmentation result set.
The present embodiment, the similarity between word and word may include the vector similarity between term vector, as cosine is similar Degree etc..Therefore, " according to the similarity between word in word in title word segmentation result set and the text word segmentation result set, from this Label of the corresponding word as the graph text information is chosen in title word segmentation result set " process may include:
Obtain in the text word segmentation result set title in the positive corresponding term vector of cliction and title word segmentation result set The corresponding term vector of word;
According to the positive corresponding term vector of cliction in the text word segmentation result set, obtain the representative word of graph text information text to Amount;
According to the corresponding term vector of heading in title word segmentation result set, similar to the vector between representing term vector Degree, label of the corresponding word as the graph text information is chosen from the title word segmentation result set.
Wherein, there are many modes for obtaining term vector, and word2vector vector training is carried out to word for example, may be used, To obtain the term vector of word.
Specifically, the vectorial phase of the term vector and representation vector of each heading in title word segmentation result set can be calculated Like degree, then, meet the heading of preset condition as tag from vector similarity is chosen.
For example, when vector similarity is cosine similarity, the word of title word segmentation result set can be traversed, is calculated each The term vector of word and the cosine similarity value for representing term vector, when the cosine similarity value is more than predetermined threshold value, it is determined that should Word is the tag of graph text information.Wherein, predetermined threshold value can be set according to actual demand, and such as 0.5.In the present embodiment, two words to The cosine similarity value of amount is bigger closer to 1, shows that the angle of two term vectors is smaller, and then show that two term vectors are more similar And the word of two term vectors is more similar.
In order to accelerate the extraction of tag and improve the accuracy of tag, the present embodiment can be selected from text word segmentation result set A certain number of positive clictions are taken, then, the representation vector of literary message text are obtained according to the positive selected works word of selection.For example, mark Extraction unit 306 is signed, can be used for choosing the target text word of preset quantity from the text word segmentation result set, and obtains mesh The corresponding term vector of positive cliction is marked, the representative term vector of graph text information text is obtained according to the corresponding term vector of target text word.
When it is implemented, above each unit can be realized as independent entity, arbitrary combination can also be carried out, is made It is realized for same or several entities, the specific implementation of above each unit can be found in the embodiment of the method for front, herein not It repeats again.
The graph text information recommendation apparatus can be specifically integrated in the equipment such as server.
From the foregoing, it will be observed that the embodiment of the present invention receives the graph text information that terminal is sent by receiving unit 301 obtains request, it should Graph text information obtains request and carries graph text information mark, then, is obtained according to the graph text information by graph text information transmission unit 302 Request sends the graph text information to the terminal and identifies corresponding target graph text information, the target picture and text are determined by determination unit 303 Whether information is to publish in instalments graph text information, is published in instalments if so, obtaining this from candidate graph text information set by publishing in instalments acquiring unit 304 Graph text information is published in instalments in the follow-up of graph text information, and sending the follow-up picture and text of publishing in instalments to the terminal from Periodical front cover information transmission unit 305 believes The Periodical front cover information of breath, to recommend this subsequently to publish in instalments graph text information.The program, which can identify, publishes in instalments graph text information, and will publish in instalments picture and text It publishes in instalments graph text information and recommends out in the follow-up of information;It may be implemented to split the graph text information comprising a large amount of pictures using the program Graph text information is published in instalments at several, this, which is published in instalments the follow-up of graph text information, after sending some to terminal and publishing in instalments graph text information connects It carries graph text information to recommend out, user is facilitated to read complete original graph text information;Therefore, the program can promote graph text information Transmission speed and graph text information displaying speed, and then improve graph text information displaying promptness and user reading body It tests.
Example IV,
Correspondingly, the embodiment of the present invention additionally provides a kind of server, as shown in figure 4, it illustrates the embodiment of the present invention The structural schematic diagram of involved server, specifically:
The server 400 may include one or more than one processing core processor 401, one or more The memory 402 of computer readable storage medium, radio frequency (Radio Frequency, RF) circuit 403, power supply 404, input are single First 405 and display unit 406 etc. components.It will be understood by those skilled in the art that server architecture shown in Fig. 4 is not The restriction to server is constituted, may include either combining certain components or different than illustrating more or fewer components Component is arranged.Wherein:
Processor 401 is the control centre of the server, utilizes each of various interfaces and the entire server of connection Part by running or execute the software program and/or module that are stored in memory 402, and calls and is stored in memory Data in 402, the various functions and processing data of execute server, to carry out integral monitoring to server.Optionally, locate Reason device 401 may include one or more processing cores;Preferably, processor 401 can integrate application processor and modulatedemodulate is mediated Manage device, wherein the main processing operation system of application processor, user interface and application program etc., modem processor is main Processing wireless communication.It is understood that above-mentioned modem processor can not also be integrated into processor 401.
Memory 402 can be used for storing software program and module.Processor 401 is stored in memory 402 by operation Software program and module, to perform various functions application and data processing.
During RF circuits 403 can be used for receiving and sending messages, signal sends and receivees, and particularly, the downlink of base station is believed After breath receives, one or the processing of more than one processor 401 are transferred to.In addition, the data for being related to uplink are sent to base station.
Server further includes the power supply 404 (such as battery) powered to all parts.Preferably, power supply can pass through power supply Management system and processor 401 are logically contiguous, to realize management charging, electric discharge and power consumption pipe by power-supply management system The functions such as reason.Power supply 404 can also include one or more direct current or AC power, recharging system, power failure The random components such as detection circuit, power supply changeover device or inverter, power supply status indicator.
The server may also include input unit 405, which can be used for receiving the number or character letter of input Breath, and generation keyboard related with user setting and function control, mouse, operating lever, optics or trace ball signal are defeated Enter.
The server may also include display unit 406, the display unit 406 can be used for showing information input by user or Be supplied to the information of user and the various graphical user interface of server, these graphical user interface can by figure, text, Icon, video and its arbitrary combination are constituted.Display unit 408 may include display panel, optionally, liquid crystal display may be used Device (LCD, Liquid Crystal Display), Organic Light Emitting Diode (OLED, Organic Light-Emitting ) etc. Diode forms configure display panel.
Specifically in the present embodiment, the processor 401 in server can be according to following instruction, by one or more The corresponding executable file of process of application program be loaded into memory 402, and be stored in by processor 401 to run Application program in reservoir 402 is as follows to realize various functions:
It receives the graph text information that terminal is sent and obtains request, which obtains request and carry graph text information mark, so Afterwards, it is obtained and is asked according to the graph text information, sending the graph text information to the terminal identifies corresponding target graph text information, and determining should Whether target graph text information is to publish in instalments graph text information, if so, obtaining this from candidate graph text information set publishes in instalments graph text information It is follow-up publish in instalments graph text information, the follow-up Periodical front cover information for publishing in instalments graph text information is sent to the terminal, to recommend this subsequently to publish in instalments Graph text information.The program, which can identify, publishes in instalments graph text information, and follow-up publish in instalments graph text information by publish in instalments graph text information and recommend Come.
For example, determine the target graph text information whether be publish in instalments graph text information, including
Obtain the title word segmentation result set of the target graph text information;
Judge in the title word segmentation result set to whether there is and the default word for publishing in instalments Keywords matching;
If so, determining that the target graph text information is to publish in instalments graph text information.
Graph text information is published in instalments for example, obtaining this from candidate graph text information set and publishing in instalments the follow-up of graph text information, including:
Obtain in candidate's graph text information set candidate graph text information author information and this publish in instalments the work of graph text information Person's information;
According to the author information of candidate's graph text information and this publish in instalments the author information of graph text information, from candidate's picture and text Information aggregate chooses corresponding candidate graph text information, obtains author's correlation candidate graph text information set;
This is chosen from author's correlation candidate graph text information set publish in instalments the follow-up of graph text information publish in instalments graph text information.
Specific operating procedure or process can refer to the detailed description of front picture and text information recommendation method embodiment.
One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment is can It is completed with instructing relevant hardware by program, which can be stored in a computer readable storage medium, storage Medium may include:Read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), disk or CD etc..
Being provided for the embodiments of the invention a kind of graph text information above recommends method, apparatus and system to carry out detailed Jie It continues, principle and implementation of the present invention are described for specific case used herein, and the explanation of above example is only It is the method and its core concept for being used to help understand the present invention;Meanwhile for those skilled in the art, according to the present invention Thought, there will be changes in the specific implementation manner and application range, in conclusion the content of the present specification should not be construed as Limitation of the present invention.

Claims (15)

1. a kind of graph text information recommends method, which is characterized in that including:
It receives the graph text information that terminal is sent and obtains request, the graph text information obtains request and carries graph text information mark;
It is obtained and is asked according to the graph text information, sending the graph text information to the terminal identifies corresponding target picture and text letter Breath;
Determine whether the target graph text information is to publish in instalments graph text information;
If so, publishing in instalments the follow-up of graph text information described in being obtained from candidate graph text information set publishes in instalments graph text information;
The follow-up Periodical front cover information for publishing in instalments graph text information is sent to the terminal, to recommend described subsequently to publish in instalments graph text information.
2. graph text information as described in claim 1 recommends method, which is characterized in that determine the target graph text information whether be Graph text information is published in instalments, including:
Obtain the title word segmentation result set of the target graph text information;
Judge in the title word segmentation result set to whether there is and the default word for publishing in instalments Keywords matching;
If so, determining that the target graph text information is to publish in instalments graph text information.
3. graph text information as described in claim 1 recommends method, which is characterized in that obtain institute from candidate graph text information set It states and publishes in instalments the follow-up of graph text information and publish in instalments graph text information, including:
Obtain the author information of candidate graph text information and the work for publishing in instalments graph text information in the candidate graph text information set Person's information;
According to the author information of the candidate graph text information and the author information for publishing in instalments graph text information, from the candidate figure Literary information aggregate chooses corresponding candidate graph text information, obtains author's correlation candidate graph text information set;
The follow-up of graph text information is published in instalments described in being chosen from author's correlation candidate graph text information set publishes in instalments graph text information.
4. graph text information as claimed in claim 3 recommends method, which is characterized in that from author's correlation candidate graph text information The follow-up of graph text information is published in instalments described in being chosen in set publishes in instalments graph text information, including:
Title in author's correlation candidate graph text information set the default candidate graph text information for publishing in instalments keyword is not included into It removes, candidate graph text information set after being removed;
Believed according to the graph text information type of candidate graph text information in candidate graph text information set after removal and the picture and text of publishing in instalments The graph text information type of breath chooses corresponding candidate graph text information from candidate graph text information set after removal, obtains type phase Close candidate graph text information set;
The follow-up of graph text information is published in instalments described in being chosen from the type correlation candidate graph text information set publishes in instalments graph text information.
5. graph text information as claimed in claim 4 recommends method, which is characterized in that from the type correlation candidate graph text information The follow-up of graph text information is published in instalments described in being chosen in set publishes in instalments graph text information, including:
Obtain the label of candidate graph text information and the mark for publishing in instalments graph text information in type correlation candidate graph text information set Label;
According to the label of the candidate graph text information and the label for publishing in instalments graph text information, believe from type correlation candidate picture and text The follow-up of graph text information is published in instalments described in being chosen in breath set publishes in instalments graph text information.
6. graph text information as described in claim 4 or 5 recommends method, which is characterized in that believe receiving the picture and text that terminal is sent Before breath obtains request, the graph text information recommendation method further includes:
Tag extraction is carried out to the graph text information in graph text information set, obtains the mark of each graph text information in graph text information set Label;
Classify to the graph text information in the graph text information set, obtains the figure of each graph text information in graph text information set Literary information type.
7. graph text information as claimed in claim 6 recommends method, which is characterized in that the graph text information in graph text information set Tag extraction is carried out, including:
Obtain the text word segmentation result set and title word segmentation result set of graph text information in graph text information set;
According to the similarity between word in word in title word segmentation result set and the text word segmentation result set, from the mark Label of the corresponding word as the graph text information is chosen in topic word segmentation result set.
8. graph text information as claimed in claim 7 recommends method, which is characterized in that according to word in title word segmentation result set, With the similarity between word in the text word segmentation result set, corresponding word is chosen from the title word segmentation result set and is made For the label of the graph text information, including:
Obtain in the text word segmentation result set heading in the positive corresponding term vector of cliction and title word segmentation result set Corresponding term vector;
According to the positive corresponding term vector of cliction in the text word segmentation result set, obtain the representative word of graph text information text to Amount;
According to the corresponding term vector of heading in title word segmentation result set, similar to the vector represented between term vector Degree, label of the corresponding word as the graph text information is chosen from the title word segmentation result set.
9. a kind of graph text information recommendation apparatus, which is characterized in that including:
Receiving unit, the graph text information for receiving terminal transmission obtain request, and the graph text information obtains request and carries picture and text Message identification;
Graph text information transmission unit is asked for being obtained according to the graph text information, and the graph text information is sent to the terminal Identify corresponding target graph text information;
Determination unit, for determining whether the target graph text information is to publish in instalments graph text information;
Acquiring unit is published in instalments, when for determining the target graph text information in the determination unit to publish in instalments graph text information, from time It selects and publishes in instalments the follow-up of graph text information described in being obtained in graph text information set and publish in instalments graph text information;
Periodical front cover information transmission unit, for sending the follow-up Periodical front cover information for publishing in instalments graph text information to the terminal, to recommend It is described subsequently to publish in instalments graph text information.
10. graph text information recommendation apparatus as claimed in claim 9, which is characterized in that the determination unit, including:
Participle obtains subelement, the title word segmentation result set for obtaining the target graph text information;
Judgment sub-unit publishes in instalments Keywords matching in the title word segmentation result set for judging with the presence or absence of with default Word;
Determination subelement, for when the judgment sub-unit is judged as YES, determining that the target graph text information is to publish in instalments picture and text Information.
11. graph text information recommendation apparatus as claimed in claim 9, which is characterized in that it is described to publish in instalments acquiring unit, including:
Author information obtains subelement, obtain in the candidate graph text information set author information of candidate graph text information and The author information for publishing in instalments graph text information;
Author chooses subelement, for according to the candidate graph text information author information and the work for publishing in instalments graph text information Person's information chooses corresponding candidate graph text information from the candidate graph text information set, obtains author's correlation candidate graph text information Set;
Acquisition of information subelement is published in instalments, for publishing in instalments picture and text letter described in the selection from author's correlation candidate graph text information set Publish in instalments graph text information in the follow-up of breath.
12. graph text information recommendation apparatus as claimed in claim 11, which is characterized in that it is described to publish in instalments acquisition of information subelement, For:
Title in author's correlation candidate graph text information set the default candidate graph text information for publishing in instalments keyword is not included into It removes, candidate graph text information set after being removed;
Believed according to the graph text information type of candidate graph text information in candidate graph text information set after removal and the picture and text of publishing in instalments The graph text information type of breath chooses corresponding candidate graph text information from candidate graph text information set after removal, obtains type phase Close candidate graph text information set;
The follow-up of graph text information is published in instalments described in being chosen from the type correlation candidate graph text information set publishes in instalments graph text information.
13. graph text information recommendation apparatus as claimed in claim 12, which is characterized in that it is described to publish in instalments acquisition of information subelement, It is specifically used for:The label of candidate graph text information and the picture and text of publishing in instalments in type correlation candidate graph text information set are obtained to believe The label of breath;According to the label of the candidate graph text information and the label for publishing in instalments graph text information, from type correlation candidate The follow-up of graph text information is published in instalments described in being chosen in graph text information set publishes in instalments graph text information.
14. graph text information recommendation apparatus as described in claim 12 or 13, which is characterized in that further include:
Tag extraction unit is used for before receiving unit receives graph text information acquisition request, to the figure in graph text information set Literary information carries out tag extraction, obtains the label of each graph text information in graph text information set;
Taxon is used for before receiving unit receives graph text information acquisition request, to the figure in the graph text information set Literary information is classified, and the graph text information type of each graph text information in graph text information set is obtained.
15. a kind of graph text information commending system, which is characterized in that including such as claim 9-14 any one of them graph text information Recommendation apparatus.
CN201710264284.9A 2017-04-21 2017-04-21 Image-text information recommendation method, device and system Active CN108304425B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710264284.9A CN108304425B (en) 2017-04-21 2017-04-21 Image-text information recommendation method, device and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710264284.9A CN108304425B (en) 2017-04-21 2017-04-21 Image-text information recommendation method, device and system

Publications (2)

Publication Number Publication Date
CN108304425A true CN108304425A (en) 2018-07-20
CN108304425B CN108304425B (en) 2021-01-08

Family

ID=62872114

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710264284.9A Active CN108304425B (en) 2017-04-21 2017-04-21 Image-text information recommendation method, device and system

Country Status (1)

Country Link
CN (1) CN108304425B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110888881A (en) * 2019-11-21 2020-03-17 望海康信(北京)科技股份公司 Picture association method and device, computer equipment and storage medium
CN113157972A (en) * 2021-04-14 2021-07-23 北京达佳互联信息技术有限公司 Recommendation method and device for video cover documents, electronic equipment and storage medium

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101976259A (en) * 2010-11-03 2011-02-16 百度在线网络技术(北京)有限公司 Method and device for recommending series documents
CN103412880A (en) * 2013-07-17 2013-11-27 百度在线网络技术(北京)有限公司 Method and device for determining implicit associated information between multimedia resources
CN105893423A (en) * 2015-12-03 2016-08-24 乐视网信息技术(北京)股份有限公司 Video recommendation method and system, terminal and server
CN105912549A (en) * 2015-12-15 2016-08-31 乐视网信息技术(北京)股份有限公司 Content recommendation method and device thereof
CN106202057A (en) * 2016-08-30 2016-12-07 东软集团股份有限公司 The recognition methods of similar news information and device
US10108674B1 (en) * 2014-08-26 2018-10-23 Twitter, Inc. Method and system for topic disambiguation and classification

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101976259A (en) * 2010-11-03 2011-02-16 百度在线网络技术(北京)有限公司 Method and device for recommending series documents
CN103412880A (en) * 2013-07-17 2013-11-27 百度在线网络技术(北京)有限公司 Method and device for determining implicit associated information between multimedia resources
US10108674B1 (en) * 2014-08-26 2018-10-23 Twitter, Inc. Method and system for topic disambiguation and classification
CN105893423A (en) * 2015-12-03 2016-08-24 乐视网信息技术(北京)股份有限公司 Video recommendation method and system, terminal and server
CN105912549A (en) * 2015-12-15 2016-08-31 乐视网信息技术(北京)股份有限公司 Content recommendation method and device thereof
CN106202057A (en) * 2016-08-30 2016-12-07 东软集团股份有限公司 The recognition methods of similar news information and device

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110888881A (en) * 2019-11-21 2020-03-17 望海康信(北京)科技股份公司 Picture association method and device, computer equipment and storage medium
CN110888881B (en) * 2019-11-21 2023-03-10 望海康信(北京)科技股份公司 Picture association method and device, computer equipment and storage medium
CN113157972A (en) * 2021-04-14 2021-07-23 北京达佳互联信息技术有限公司 Recommendation method and device for video cover documents, electronic equipment and storage medium
CN113157972B (en) * 2021-04-14 2023-09-19 北京达佳互联信息技术有限公司 Recommendation method and device for video cover document, electronic equipment and storage medium

Also Published As

Publication number Publication date
CN108304425B (en) 2021-01-08

Similar Documents

Publication Publication Date Title
US9158846B2 (en) Entity detection and extraction for entity cards
CN102982178B (en) A kind of image searching method, device and system
WO2017133343A1 (en) Picture processing method and apparatus, and electronic device
US20110307482A1 (en) Search result driven query intent identification
CN103136228A (en) Image search method and image search device
CN103942279A (en) Method and device for showing search result
CN109242044A (en) Training method, device, storage medium and the electronic equipment of vehicle and goods matching model
CN104077290B (en) A kind of method and apparatus for generating promoted account
CN109462632B (en) Page jump method, page jump device, computer equipment and computer readable storage medium
US20230089850A1 (en) Real-time product environmental impact scoring
CN109635198A (en) The method, apparatus of presentation user's search result, medium and electronic equipment on merchandise display platform
CN106095738A (en) Recommendation tables single slice
CN110163703A (en) A kind of disaggregated model method for building up, official documents and correspondence method for pushing and server
US20150348131A1 (en) Method, system and recording medium for providing image using metadata of image file
CN111695041B (en) Method and device for recommending information
CN114154013A (en) Video recommendation method, device, equipment and storage medium
CN112818111A (en) Document recommendation method and device, electronic equipment and medium
CN106682146B (en) Method and system for retrieving scenic spot evaluation according to keywords
CN110059172B (en) Method and device for recommending answers based on natural language understanding
CN108304425A (en) A kind of graph text information recommends method, apparatus and system
CN113037925B (en) Information processing method, information processing apparatus, electronic device, and readable storage medium
US10891580B2 (en) Service design assistance system and service design assistance method
CN116467607B (en) Information matching method and storage medium
CN112306606A (en) Method and equipment for presenting control style
US20170161389A1 (en) Determining an object referenced within informal online communications

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
TR01 Transfer of patent right

Effective date of registration: 20221114

Address after: 1402, Floor 14, Block A, Haina Baichuan Headquarters Building, No. 6, Baoxing Road, Haibin Community, Xin'an Street, Bao'an District, Shenzhen, Guangdong 518,101

Patentee after: Shenzhen Yayue Technology Co.,Ltd.

Address before: 518000 Tencent Building, No. 1 High-tech Zone, Nanshan District, Shenzhen City, Guangdong Province, 35 Floors

Patentee before: TENCENT TECHNOLOGY (SHENZHEN) Co.,Ltd.

TR01 Transfer of patent right