CN104077327A - Core word importance recognition method and equipment and search result sorting method and equipment - Google Patents
Core word importance recognition method and equipment and search result sorting method and equipment Download PDFInfo
- Publication number
- CN104077327A CN104077327A CN201310109430.2A CN201310109430A CN104077327A CN 104077327 A CN104077327 A CN 104077327A CN 201310109430 A CN201310109430 A CN 201310109430A CN 104077327 A CN104077327 A CN 104077327A
- Authority
- CN
- China
- Prior art keywords
- information
- word
- core
- core word
- initial weight
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to a core word importance recognition method and equipment and a search result sorting method and equipment. The core word importance recognition method is used for recognizing importance of core words in information released by an information release user, and is characterized by including: determining multiple core words in the released information; according to features of the released information, giving a corresponding initial weight value to each of the core words; according to a historical behavior log of the information release user, regulating the initial weight value corresponding to each core word so as to obtain a corresponding final weight value. Recognition accuracy of the importance of the core words can be improved, so that searching and sorting accuracy of the related released information is improved.
Description
Technical field
The application relates to field of computer technology, relates more specifically to the recognition methods of core word importance and equipment and search result ordering method and equipment for the importance of identifying information issue user's the core word that releases news.
Background technology
Although the content is here set forth under background technology title, the inventor's discovery and design have wherein also been comprised, so prior art should be considered as completely.
Along with the fast development of internet, by the network platform, release news and search information has become the daily one way of life mode of people.Therefore, search engine technique is also in constantly innovation and development, to meet people's expectation and demand.
In this manual, " information issue user " refers to the user who releases news in the network platform, and " information search user " refers to the user who searches for information in the network platform.
Traditionally, in search engine, conventional retrieval mode is following information retrieval model, the information (being called " releasing news ") of wherein issuing for information issue user is set up index database, then according to information search user's search word, by literal hitting with matching process, determine the correlativity releasing news with search word, and show release news relevant to search word according to relevance ranking.
Yet, in such method, only literal hit and mate may cause recalling a large amount of non-search word demand informations.For example, in the network platform, exist two to release news: (A) to supply trendy Nokia smart mobile phone; (B) supply Nokia battery of mobile phone.When information search user utilizes " mobile phone " to search for as search word, the retrieval logic of general search engine adopts literal hitting and matching process, then determine these two release news all relevant to this search word, so these two information all can be called back.Yet in fact only having (A) is the demand that meets information search user, (B) is irrelevant information for this information search user, such search accuracy is lower, can not meet consumers' demand well.
For solving such problem, at present search engine generally releases news to identify this core presentive word or core product word (referred to as core word) in releasing news by analysis, and using it as search engine, recall the correlation calculations foundation of result, rather than by whole piece release news with information search user's search word carry out literal hit and mate determine correlativity, can reduce to a certain extent literal coupling thus and the recalling of the incoherent information of meaning.
In such method, the common way of identification core word is, releasing news of user of information issue carried out to word segmentation processing and part-of-speech tagging, according to pre-prepd part of speech or attribute dictionary matching, to releasing news, mark, or by machine learning method automatic marking part of speech or attribute, then, according to such as the pre-defined rule as core word using the word that is labeled as noun, identify the core word that this releases news.
For those, describe lack of standardization or descriptor is more releases news, may identify a plurality of core words by this method.For example, release news as " supplying trendy notebook computer, 500G hard disk, 4G internal memory, 15 cun of liquid crystal display ", by said method, " notebook computer ", " hard disk ", " internal memory " and " liquid crystal display " all can be identified as to core word.Under these circumstances; conventionally meeting basis is such as TF-IDF(term frequency – inverse document frequency; word frequency-anti-document frequency) and so on method is identified the importance of these core words in this releases news; then according to the importance of core word, determine the correlativity that this releases news with search word, and show release news relevant to search word according to relevance ranking.
Yet the inventor finds, the in the situation that of this plurality of core word, that is, lack of standardization or descriptor is described more in the situation that releasing news, search accuracy is often not high.Therefore, expect that a kind of technology that can improve search accuracy overcomes this defect.
Summary of the invention
The inventor notices, in the core word importance identification of classic method, the importance of word is along with the increase that is directly proportional of number of times that it occurs hereof, but the decline that simultaneously can be inversely proportional to along with the frequency that it occurs in corpus.Wherein only utilized information issue user's the descriptor that releases news own.Such identification is not accurate enough, and mistake appears in the correlation calculations that causes subsequent searches engine to recall result, thereby causes the accuracy of relevance ranking not high, can not meet consumers' demand well.
Therefore, the application object is to provide a kind of technology that the importance of information issue user's the middle core word that releases news is identified.
Another object of the application is to provide a kind of technology that the relevant Search Results releasing news is sorted.
According to the embodiment of the application aspect, the recognition methods of a kind of core word importance is provided, for identifying information, issue the importance of user's the core word that releases news, it is characterized in that, comprising: determine a plurality of core words in releasing news; According to the feature releasing news, for each core word in a plurality of core words is given corresponding initial weight value; And according to information issue user's historical behavior daily record, adjust the corresponding initial weight value of each core word, to obtain corresponding final weighted value.
Embodiment according to the application aspect, also provides a kind of search result ordering method, it is characterized in that, comprising: the search word that receives information inquiry user input; The importance of the middle core word that releases news based on information issue user, determines the correlativity releasing news with search word; And according to correlativity, to releasing news, sort and show, wherein, the importance of described core word is identified by following steps: a plurality of core words in releasing news described in determining; According to the described feature releasing news, for each core word in described a plurality of core words is given corresponding initial weight value; And according to information issue user's historical behavior daily record, adjust the described corresponding initial weight value of each core word, to obtain corresponding final weighted value.
According to the application's embodiment on the other hand, a kind of core word importance identification equipment is provided, for identifying information, issues the importance of user's the core word that releases news, it is characterized in that, comprise: core word determining device, for determining a plurality of core words that release news; Valuator device, the feature releasing news for basis, for each core word in a plurality of core words is given corresponding initial weight value; And adjusting gear, for according to information issue user's historical behavior daily record, adjust the corresponding initial weight value of each core word, to obtain corresponding final weighted value.
According to the application's embodiment on the other hand, a kind of search results ranking equipment is also provided, it is characterized in that, comprising: search word receiving trap, for receiving the search word of information inquiry user input; Correlativity determining device, for the importance of the core word that releases news based on information issue user, determines the correlativity releasing news with search word; And sequence and display device, for according to correlativity, to releasing news, sort and show, wherein, the importance of described core word is identified by following steps: a plurality of core words in releasing news described in determining; According to the described feature releasing news, for each core word in described a plurality of core words is given corresponding initial weight value; And according to information issue user's historical behavior daily record, adjust the described corresponding initial weight value of each core word, to obtain corresponding final weighted value.
Compared with prior art, according to the application's technical scheme, not only according to information issue user's the feature that releases news own, and other suitable supplementary combination such as information issue user's historical behavior daily record, identify the importance of core word in this releases news, thus the recognition accuracy that can improve the importance of the core word in releasing news.Correspondingly, can improve search accuracy, that is, improve the sequence accuracy that release news relevant to search word, thereby meet consumers' demand better.
Accompanying drawing explanation
Accompanying drawing described herein is used to provide further understanding of the present application, forms the application's a part, and the application's schematic description and description is used for explaining the application, does not form the improper restriction to the application.In the accompanying drawings:
Fig. 1 illustrates according to the process flow diagram of the core word importance recognition methods of the importance of the core word that releases news for identifying information issue user of an embodiment of the application;
Fig. 2 illustrates according to the process flow diagram of the search result ordering method of an embodiment of the application;
Fig. 3 illustrates according to the schematic block diagram of the core word importance identification equipment of the importance of the core word that releases news for identifying information issue user of an embodiment of the application; And
Fig. 4 illustrates according to the schematic block diagram of the search results ranking equipment of an embodiment of the application.
Embodiment
As mentioned above, the inventor notices, itself the feature of releasing news that only merely combining information releases news is identified the importance of core word, and such recognition accuracy is not high.So the inventor expects, can optimize in conjunction with the suitable supplementary the own feature that releases news except information issue user the result of core word importance identification.
The application's main thought is, except information issue user's the feature that releases news own, the historical behavior daily record of considering combining information issue user carrys out the importance of identifying information issue user's the middle core word that releases news, to improve the accuracy of importance identification.
The inventor notices, information inquiry user's feedback information is also an available important and high-quality information, can optimize the result of core word importance identification by such feedback information.Be further noted that, information issue user's personal information is also can be for promoting the good foundation of importance recognition accuracy.
For making the application's object, technical scheme and advantage clearer, below in conjunction with drawings and the specific embodiments, the application is described in further detail.
According to the embodiment of an aspect of the application, provide a kind of core word importance recognition methods of importance of the core word that releases news for identifying information issue user.
Fig. 1 illustrates according to the process flow diagram of the core word importance recognition methods of the importance of the core word that releases news for identifying information issue user of an embodiment of the application.
As shown in Figure 1, at step S110 place, determine a plurality of core words in releasing news.
To releasing news of user of information issue, can carry out the analysis such as word segmentation processing and part-of-speech tagging, and determine the core word in releasing news according to pre-defined rule.
In a specific embodiment, for example, can release news and carry out word segmentation processing and part-of-speech tagging this, and the core word in the word that is labeled as noun part of speech or product word can being defined as releasing news.
In one releases news, can there is one or more nouns or product word.The application is mainly for releasing news and comprise the situation of a plurality of core words in one.
Here it is pointed out that and can determine the core word in releasing news by any desired manner of known in the art or following exploitation, and be not limited to mode listed above.
Next, at step S120 place, the feature according to core word in releasing news, for each core word in a plurality of core words is given corresponding initial weight value Score_initial.
Wherein, this initial weight value can be for tentatively identifying the importance of this core word in this releases news.
Wherein, release news and can comprise the descriptor of title, attribute and/or the details and so on that are published object.
The feature of core word in releasing news can comprise that core word is at the frequency of middle appearance that releases news (number of times) and/or position feature.
In a specific embodiment, for example, core word is more at the number of times of the middle appearance that releases news, and weighted value Score_initial is higher.In addition, for example, if core word appears at during title describes, weighted value Score_initial is high, and if core word only appears at during details describe, weighted value Score_initial is low.These features can be used separately also and can be combined with.This point is well to realize by any desired manner of known in the art or following exploitation, repeats no more here.
As previously mentioned, the inventor notices just, in existing scheme, only according to the own feature that releases news, give each core word weighted value, by this weighted value, identify the importance of a plurality of core words, such importance recognition accuracy is not high, cause in information search this to release news not high with the accuracy of the correlation calculations of user input query word, so expected adjusting in conjunction with other suitable supplementary the weights of importance value of this middle core word that releases news, make to improve importance recognition accuracy, be conducive to promote in search the accuracy that this releases news with the correlation calculations of user input query word.
According to the application's embodiment, can be according to issuing user's historical behavior daily record such as information, one or more in the supplementary information inquiry user's feedback information, information issue user's personal information be adjusted the initial weight value of core word, thereby improve the accuracy of identification core word importance, as integrating step S130 is described below.
At step S130 place, according to information issue user's historical behavior daily record, adjust the corresponding initial weight value of each core word Score_initial
i, to obtain corresponding final weighted value Score_final
i.
In a specific embodiment, the some core word i in an information of information issue user issue, calculate the number of times Count_key that this core word occurs in this information issue user's historical behavior daily record
iand the number of times sum ∑ Count_key that in this information, each core word occurs in this information issue user's historical behavior daily record
iratio Score_key
i, i.e. Score_key
i=Count_key
i/ ∑ Count_key
i, i represents i core word in the information of an issue.
Wherein, information issue user's historical behavior daily record specifically can comprise information issue user's keyword purchase daily record.The keyword that information issue user buys can comprise participle and minute word combination, and a minute word combination is combined by a plurality of participles.
In one embodiment, the final weighted value Score_final of each core word
ican be Score_key
iwith the weighted sum of initial weight value, shown in (2):
Score_final
i=w
5*(w
7*Score_key
i)+w
6*Score_initial
i (2)
Wherein, w
5, w
6and w
7can be the experience weights that draw according to experimental result in experiment, they can be the arbitrary value between 0-1.
Described above is according to information issue user's historical behavior daily record, to identify the importance of core word.In fact, can also be further according to other appropriate information of information issue user side, improve recognition accuracy.
In one embodiment, the personal information that can issue user according to information is adjusted the corresponding initial weight value of each core word, to obtain corresponding final weighted value.
According to the application's embodiment, personal information at least comprises at least one in individual label, summary and regional information.This personal information can obtain from information is issued user's log-on message.For example, log-on message can comprise individual label information such as title, the summary info such as remarks, the regional information such as address etc.
In a specific embodiment, the initial weight value that the frequency (number of times) that can occur in above-mentioned personal information according to core word or position adjustment are given core word.For example, core word occurs manyly in individual label information, summary info, regional information, and weighted value can be higher.This situation is similar to the situation while considering to release news feature own.According to content disclosed herein, those skilled in the art can easily realize this point, therefore repeat no more here.
Described above be according to information issue user's historical behavior daily record and/or information issue user's personal information identify release news in the importance of each core word, can think it is to adjust the initial weight value of giving core word by itself the feature of releasing news according to the information of information issue user side above.In fact, can also adjust this initial weight value according to the appropriate information of information inquiry user side.
In one embodiment, can carry out the corresponding initial weight value of each core word Score_initial determining in set-up procedure S110 according to information inquiry user's feedback information
i, to obtain corresponding final weighted value Score_final
i.
Wherein, information inquiry user's feedback information at least comprises information inquiry user's inquiry and click information, Transaction Information and evaluates at least one in behavioural information.These information inquiries user's feedback information, inquiry and click information, click subsequent transaction information and evaluation behavioural information such as information inquiry user, can obtain by network log.Here it will be appreciated that, the weighted value that dissimilar feedback information adjustment that can combining information inquiring user utilizes information issue user's feature to draw, thereby improve the recognition accuracy of core word importance, and then can improve the relevant search accuracy releasing news.
In a specific embodiment, can and click historical information and adjust the corresponding initial weight value of each core word Score_initial according to information inquiry user's inquiry
i, to obtain corresponding final weighted value Score_final
i.For example, for each core word, the number of times Count_show that can occur in the Query Result in certain hour section (such as 100 days) in network log according to this core word
iwith each the number of times sum ∑ Count_show occurring in this Query Result in a plurality of core words
iratio Score_show
i=Count_show
i/ ∑ Count_show
iand the number of times Count_click that in interior clicked Query Result of this time period (such as 100 days), this core word occurs
iwith the number of times sum ∑ Count_click that in clicked Query Result, each core word occurs
iratio Score_click
i=Count_click
i/ ∑ Count_click
i, adjust initial weight value Score_initial
ithereby obtain final weighted value Score_final
i, i represents an i core word in releasing news.In one embodiment, the final weighted value of each core word can be Score_show
i, Score_click
iand Score_initial
iweighted sum, shown in (1):
Score_final
i=w
1*(w
3*Score_show
i+w
4*Score_click
i)+w
2*Score_initial
i (1)
Wherein, w
1, w
2, w
3and w
4can be according to experimental result, to preset in experiment, they can be the arbitrary value between 0-1.
Described above is to adjust the corresponding initial weight value of each core word according to information inquiry user's inquiry and click historical information.In a similar fashion, equally can be according to clicking subsequent transaction information or evaluating behavioural information and adjust the corresponding initial weight value of each core word, also can adjust the corresponding initial weight value of each core word according to information inquiry user's inquiry and click information, click subsequent transaction information, the combination in any evaluated between behavioural information three.Those skilled in the art can realize these schemes according to disclosed content above, therefore for for purpose of brevity, about they implementations separately, repeat no more here.
Describe in the above embodiments be only combining information issue user's historical behavior daily record or only combining information issue user personal information or only the feedback information of combining information inquiring user identify the importance of core word, the recognition accuracy of the core word importance that releases news be can improve thus, thereby the relevant search releasing news and sequence accuracy improved.It should be understood that, the application is not limited to above-described embodiment, but can identify according to the combination in any in above-mentioned information the importance of core word, can further improve like this recognition accuracy and the relevant search accuracy releasing news of core word importance.
For example, in another embodiment, can the inquiry of combining information inquiring user and the feedback information of click and information issue user's historical behavior daily record the two adjust the corresponding initial weight value of each core word Score_initial
i, to obtain corresponding final weighted value Score_final
i.Shown in (3):
Score_final
i=w
1'*(w
3'*(w
5'*Score_show
i+w
6'*Score_click
i)+w
4'*(Score_key
i))+w
2'*Score_initial
i (3)
Wherein, w
1', w
2', w
3', w
4', w
5' and w
6' can be the experience weights that draw according to experimental result in experiment, they can be the arbitrary value between 0-1.
So far, by basis release news itself feature and in conjunction with supplementary, obtained the final weighted value of each core word, thereby identify the importance of each core word of a plurality of core words in releasing news, can obviously improve thus the recognition accuracy of core word importance.
When current inquiring user is inputted a certain search word, by this search word, search one or more information, according to the final weighted value of each core word in every information, calculate the correlativity of every information and this search word, and according to described correlativity calculation result to described information sorting.
In the embodiment of the present application, when searching for according to the search word of information inquiry user input, can improve the accuracy in correlation calculations and irrelevant information filtration, thereby the accuracy that improves the relevant search results ranking releasing news is described this point in detail below in conjunction with Fig. 2.
Fig. 2 illustrates according to the process flow diagram of the search result ordering method of an embodiment of the application.
As shown in Figure 2, at step S210 place, receive the search word of information inquiry user input.
In one embodiment, can analytical information inquiring user the search word of input, to find out the core word information in this search word.Generally speaking, search word is shorter character string, can well identify core word information wherein, for follow-up correlation calculations by this area common method or with the similar approach that integrating step S110 describes.
Obviously it will be appreciated that, the application is not limited to above-described embodiment, also can not find out the core word information in search word, but directly uses search word to carry out the correlation calculations of subsequent step S220.
Next, at step S220 place, the importance of the middle core word that releases news based on information issue user, determines the correlativity releasing news with search word.
Wherein the importance of information issue user's the middle core word that releases news is to obtain by the method for the application's described in conjunction with Figure 1 identification core word importance above, and its details can, with reference to description above, repeat no more here.
In one embodiment, the core word information in the search word receiving can be contrasted in the core word in predetermined releasing news and step S210, determine the correlativity releasing news with search word.
Particularly, if the high core word of final weighted value in releasing news and the core word information matches of search word determine that this releases news higher with search word correlativity.If the core word that the final weighted value in releasing news is low and the core word information matches of search word, determine that this releases news lower with search word correlativity.If the core word in releasing news does not mate with the core word information of search word, determine that this releases news irrelevant with search word.
Next, at step S230 place, according to the definite correlativity in step S220 place, relevant releasing news sorted and shown.
Namely, can be according to releasing news of determining in step S220 above the correlativity with search word, release news relevant to search word sorted and shown.Particularly, can be according to the height of correlativity, before higher the releasing news of the correlativity with search word is presented at, after being presented at lower the releasing news of correlativity of search word, and do not show with irrelevant the releasing news of search word.
In other embodiments, can, according to the final weighted value of each core word in obtained above releasing news, to a plurality of core words in releasing news, carry out importance ranking.In the application scenarioss such as the search of having relatively high expectations for correlativity or information classification, can be only at search word, just determine that this releases news during to core word coupling in the middle importance ranking first that releases news to release news for relevant.
Pass through said method, according to not only identify the importance of the middle core word that releases news in conjunction with release news feature own but also combining information issue user's historical behavior daily record, when searching for according to the search word of information inquiry user input, can improve the accuracy of the correlation calculations that releases news, thereby improve the relevant sequence accuracy releasing news, user-friendly and promote user's use impression.
Similar with the core word importance recognition methods of the importance of the above-mentioned core word that releases news for identifying information issue user, the embodiment of the present application also provides the core word importance identification equipment for the importance of identifying information issue user's the core word that releases news.
Fig. 3 illustrates according to the schematic block diagram of the core word importance identification equipment 300 of the importance of the core word that releases news for identifying information issue user of an embodiment of the application.
As shown in Figure 3, equipment 300 can comprise core word determining device 310, valuator device 320 and adjusting gear 330.
Particularly, core word determining device 310 can be for determining a plurality of core words in releasing news.Valuator device 320 can be for giving corresponding initial weight value according to the feature that release news for each core word in a plurality of core words.Adjusting gear 330 can be for adjusting the corresponding initial weight value of each core word to obtain corresponding final weighted value according to information issue user's historical behavior daily record.
The equipment of the importance of the core word that releases news for identifying information issue user by the embodiment of the present application, compared to existing technologies, can obviously improve the accuracy of identification core word importance.
The core word importance identification equipment of the importance of the core word that releases news for identifying information issue user described above is corresponding with the processing of the core word importance recognition methods of the importance of the core word that releases news for identifying information issue user of describing before, therefore, about more detailed ins and outs, can, referring to the method for describing before, repeat no more here.
On the other hand, similar with mentioned above searching results sort method, the embodiment of the present application also provides search results ranking equipment, below in conjunction with Fig. 4, describes in detail.
Fig. 4 illustrates according to the schematic block diagram of the search results ranking equipment 400 of an embodiment of the application.
As shown in Figure 4, equipment 400 can comprise search word receiving trap 410, correlativity determining device 420 and sequence and display device 430.
Particularly, search word receiving trap 410 can be for receiving the search word of information inquiry user input.Correlativity determining device 420 can be determined the correlativity releasing news with search word for the importance of the middle core word that releases news based on information issue user.Wherein the importance of the core word in the releasing news of information issue user is that the method for the identification core word importance of the application by describing in conjunction with Fig. 1 above obtains.Sequence and display device 430 can be for sorting to releasing news according to correlativity and showing.
Similarly, by the search results ranking equipment of the embodiment of the present application, the accuracy that can improve the correlation calculations that releases news, thus improve the relevant sequence accuracy releasing news, user-friendly and promote user's use impression.
Search results ranking equipment described above is corresponding with the processing of the search result ordering method of describing before, therefore, about more detailed ins and outs, can, referring to the method for describing before, repeat no more here.
The embodiment that it will be understood by those skilled in the art that the application can be provided as method, system or computer program.Therefore, the application can adopt complete hardware implementation example, implement software example or in conjunction with the form of the embodiment of software and hardware aspect completely.And the application can adopt the form that wherein includes the upper computer program of implementing of computer-usable storage medium (including but not limited to magnetic disk memory, CD-ROM, optical memory etc.) of computer usable program code one or more.
The embodiment that the foregoing is only the application, is not limited to the application.To those skilled in the art, the application can have various modifications and variations.Any modification of doing within all spirit in the application and principle, be equal to replacement, improvement etc., within all should being included in the application's claim scope.
Claims (16)
1. a core word importance recognition methods, the importance for identifying information issue user's the core word that releases news, is characterized in that, comprising:
A plurality of core words in releasing news described in determining;
According to the described feature releasing news, for each core word in described a plurality of core words is given corresponding initial weight value; And
According to described information issue user's historical behavior daily record, adjust the described corresponding initial weight value of each core word, to obtain corresponding final weighted value.
2. method according to claim 1, is characterized in that, also comprises:
According to information inquiry user's feedback information, adjust the described corresponding initial weight value of each core word, to obtain corresponding final weighted value.
3. method according to claim 2, is characterized in that, described information inquiry user's feedback information at least comprises described information inquiry user's inquiry and click information, click subsequent transaction information and evaluates at least one in behavioural information.
4. method according to claim 1 and 2, is characterized in that, also comprises:
According to described information issue user's personal information, adjust the described corresponding initial weight value of each core word, to obtain corresponding final weighted value, described personal information at least comprises at least one in individual label, summary and regional information.
5. method according to claim 1, is characterized in that, the step of described a plurality of core words in releasing news described in determining comprises:
Described releasing news carried out to word segmentation processing and part-of-speech tagging; And
Core word in releasing news described in determining according to pre-defined rule, described pre-defined rule is the core word in releasing news described in the word that is labeled as noun part of speech or product word is defined as.
6. method according to claim 1, is characterized in that, describedly according to the corresponding initial weight value of each core word described in information issue user's historical behavior daily record adjustment, to obtain the step of corresponding final weighted value, comprises:
Calculate the ratio of the number of times sum that in number of times that each core word occurs in this information issue user's historical behavior daily record and this information, each core word occurs in this information issue user's historical behavior daily record;
Thereby according to described ratio, adjust described initial weight value and obtain final weighted value.
7. method according to claim 2, is characterized in that, describedly according to the corresponding initial weight value of each core word described in information inquiry user's feedback information adjustment, to obtain the step of corresponding final weighted value, comprises:
Ratio according to the number of times sum of the number of times of each appearance in a plurality of core words described in Query Result clicked in the ratio of the number of times sum of each appearance in the number of times of each appearance in a plurality of core words described in the Query Result in certain hour section and described a plurality of core word and certain hour section and each appearance in described a plurality of core word, obtains final weighted value thereby adjust described initial weight value.
8. a search result ordering method, is characterized in that, comprising:
Receive the search word of information inquiry user input;
The importance of the middle core word that releases news based on information issue user, the correlativity releasing news described in determining with described search word; And
According to described correlativity, described releasing news sorted and shown,
Wherein, the importance of described core word is identified by following steps:
A plurality of core words in releasing news described in determining;
According to the described feature releasing news, for each core word in described a plurality of core words is given corresponding initial weight value; And
According to information issue user's historical behavior daily record, adjust the described corresponding initial weight value of each core word, to obtain corresponding final weighted value.
9. a core word importance identification equipment, the importance for identifying information issue user's the core word that releases news, is characterized in that, comprising:
Core word determining device, for a plurality of core words that release news described in determining;
Valuator device, for the feature releasing news described in basis, for each core word in described a plurality of core words is given corresponding initial weight value; And
Adjusting gear, for according to described information issue user's historical behavior daily record, adjusts the described corresponding initial weight value of each core word, to obtain corresponding final weighted value.
10. equipment according to claim 9, is characterized in that, described adjusting gear is also adjusted the described corresponding initial weight value of each core word according to information inquiry user's feedback information, to obtain corresponding final weighted value.
11. equipment according to claim 10, is characterized in that, described information inquiry user's feedback information at least comprises described information inquiry user's inquiry and click information, click subsequent transaction information and evaluates at least one in behavioural information.
12. according to the equipment described in claim 9 or 10, it is characterized in that, described adjusting gear also according to the corresponding initial weight value of each core word described in described information issue user's personal information adjustment to obtain corresponding final weighted value, described personal information at least comprises at least one in individual label, summary and regional information.
13. equipment according to claim 9, it is characterized in that, described core word determining device to described release news carry out word segmentation processing and part-of-speech tagging and according to pre-defined rule, determine described in core word in releasing news, described pre-defined rule is the core word in releasing news described in the word that is labeled as noun part of speech or product word is defined as.
14. equipment according to claim 9, it is characterized in that, described adjusting gear calculates the ratio of the number of times sum that in number of times that each core word occurs in this information issue user's historical behavior daily record and this information, each core word occurs in this information issue user's historical behavior daily record, thereby and according to described ratio, adjusts described initial weight value and obtain final weighted value.
15. equipment according to claim 10, it is characterized in that, described adjusting gear, according to the ratio of the number of times sum of the number of times of each appearance in a plurality of core words described in Query Result clicked in the ratio of the number of times of each appearance in the number of times of each appearance in a plurality of core words described in the Query Result in certain hour section and described a plurality of core word and certain hour section and each appearance in described a plurality of core word, obtains final weighted value thereby adjust described initial weight value.
16. 1 kinds of search results ranking equipment, is characterized in that, comprising:
Search word receiving trap, for receiving the search word of information inquiry user input;
Correlativity determining device, for the importance of the core word that releases news based on information issue user, the correlativity releasing news described in determining with described search word; And
Sequence and display device, for according to described correlativity, sort and show described releasing news,
Wherein, the importance of described core word is identified by following steps:
A plurality of core words in releasing news described in determining;
According to the described feature releasing news, for each core word in described a plurality of core words is given corresponding initial weight value; And
According to information issue user's historical behavior daily record, adjust the described corresponding initial weight value of each core word, to obtain corresponding final weighted value.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310109430.2A CN104077327B (en) | 2013-03-29 | 2013-03-29 | The recognition methods of core word importance and equipment and search result ordering method and equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310109430.2A CN104077327B (en) | 2013-03-29 | 2013-03-29 | The recognition methods of core word importance and equipment and search result ordering method and equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104077327A true CN104077327A (en) | 2014-10-01 |
CN104077327B CN104077327B (en) | 2018-01-19 |
Family
ID=51598586
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310109430.2A Active CN104077327B (en) | 2013-03-29 | 2013-03-29 | The recognition methods of core word importance and equipment and search result ordering method and equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104077327B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105205045A (en) * | 2015-09-21 | 2015-12-30 | 上海智臻智能网络科技股份有限公司 | Semantic model method for intelligent interaction |
CN105893397A (en) * | 2015-06-30 | 2016-08-24 | 北京爱奇艺科技有限公司 | Video recommendation method and apparatus |
CN107688606A (en) * | 2017-07-26 | 2018-02-13 | 北京三快在线科技有限公司 | The acquisition methods and device of a kind of recommendation information, electronic equipment |
CN111949697A (en) * | 2020-07-09 | 2020-11-17 | 厦门美柚股份有限公司 | Data processing method, device, terminal and medium based on search engine |
CN107818781B (en) * | 2017-09-11 | 2021-08-10 | 远光软件股份有限公司 | Intelligent interaction method, equipment and storage medium |
CN113761110A (en) * | 2020-06-28 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Information issuing method, device, equipment and storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101334796A (en) * | 2008-02-29 | 2008-12-31 | 浙江师范大学 | Personalized and synergistic integration network multimedia search and enquiry method |
CN102289436A (en) * | 2010-06-18 | 2011-12-21 | 阿里巴巴集团控股有限公司 | Method and device for determining weighted value of search term and method and device for generating search results |
US8145618B1 (en) * | 2004-02-26 | 2012-03-27 | Google Inc. | System and method for determining a composite score for categorized search results |
CN102446174A (en) * | 2010-10-09 | 2012-05-09 | 百度在线网络技术(北京)有限公司 | Method for determining weights of key sub-words in network equipment and equipment adopting same |
-
2013
- 2013-03-29 CN CN201310109430.2A patent/CN104077327B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8145618B1 (en) * | 2004-02-26 | 2012-03-27 | Google Inc. | System and method for determining a composite score for categorized search results |
CN101334796A (en) * | 2008-02-29 | 2008-12-31 | 浙江师范大学 | Personalized and synergistic integration network multimedia search and enquiry method |
CN102289436A (en) * | 2010-06-18 | 2011-12-21 | 阿里巴巴集团控股有限公司 | Method and device for determining weighted value of search term and method and device for generating search results |
CN102446174A (en) * | 2010-10-09 | 2012-05-09 | 百度在线网络技术(北京)有限公司 | Method for determining weights of key sub-words in network equipment and equipment adopting same |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105893397A (en) * | 2015-06-30 | 2016-08-24 | 北京爱奇艺科技有限公司 | Video recommendation method and apparatus |
CN105893397B (en) * | 2015-06-30 | 2019-03-15 | 北京爱奇艺科技有限公司 | A kind of video recommendation method and device |
CN105205045A (en) * | 2015-09-21 | 2015-12-30 | 上海智臻智能网络科技股份有限公司 | Semantic model method for intelligent interaction |
CN107688606A (en) * | 2017-07-26 | 2018-02-13 | 北京三快在线科技有限公司 | The acquisition methods and device of a kind of recommendation information, electronic equipment |
WO2019019554A1 (en) * | 2017-07-26 | 2019-01-31 | 北京三快在线科技有限公司 | Method and apparatus for obtaining recommendation information, and electronic device |
CN107818781B (en) * | 2017-09-11 | 2021-08-10 | 远光软件股份有限公司 | Intelligent interaction method, equipment and storage medium |
CN113761110A (en) * | 2020-06-28 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Information issuing method, device, equipment and storage medium |
CN111949697A (en) * | 2020-07-09 | 2020-11-17 | 厦门美柚股份有限公司 | Data processing method, device, terminal and medium based on search engine |
CN111949697B (en) * | 2020-07-09 | 2022-08-16 | 厦门美柚股份有限公司 | Data processing method, device, terminal and medium based on search engine |
Also Published As
Publication number | Publication date |
---|---|
CN104077327B (en) | 2018-01-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107818781B (en) | Intelligent interaction method, equipment and storage medium | |
WO2018050022A1 (en) | Application program recommendation method, and server | |
CN106575503B (en) | Method and system for session context modeling for dialog understanding systems | |
WO2010022655A1 (en) | A searching method and system | |
CN106601237B (en) | Interactive voice response system and voice recognition method thereof | |
CN101241512B (en) | Search method for redefining enquiry word and device therefor | |
CN102929873B (en) | Method and device for extracting searching value terms based on context search | |
CN102402619B (en) | Search method and device | |
CN107180093B (en) | Information searching method and device and timeliness query word identification method and device | |
CN103106287B (en) | A kind of processing method and system of user search sentence | |
CN104077327A (en) | Core word importance recognition method and equipment and search result sorting method and equipment | |
US20110145348A1 (en) | Systems and methods for identifying terms relevant to web pages using social network messages | |
CN104199822A (en) | Method and system for identifying demand classification corresponding to searching | |
CN103605665A (en) | Keyword based evaluation expert intelligent search and recommendation method | |
CN101551806A (en) | Personalized website navigation method and system | |
CN101641697A (en) | Related search queries for a webpage and their applications | |
CA2536265A1 (en) | System and method for processing a query | |
CN111090771B (en) | Song searching method, device and computer storage medium | |
CN102456054B (en) | A kind of searching method and system | |
WO2011054245A1 (en) | Mobile search method, device and system | |
TWI662495B (en) | Processing method, device and system for promotion information | |
CN102637179B (en) | Method and device for determining lexical item weighting functions and searching based on functions | |
CN108038099B (en) | Low-frequency keyword identification method based on word clustering | |
CN109145110A (en) | Information classification processing, tag queries method and apparatus based on label | |
CN103365915A (en) | Search result ranking method based on search engine and database query system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |