CN108170664B - Key word expansion method and device based on key words - Google Patents

Key word expansion method and device based on key words Download PDF

Info

Publication number
CN108170664B
CN108170664B CN201711227953.1A CN201711227953A CN108170664B CN 108170664 B CN108170664 B CN 108170664B CN 201711227953 A CN201711227953 A CN 201711227953A CN 108170664 B CN108170664 B CN 108170664B
Authority
CN
China
Prior art keywords
keyword
keywords
app
key
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.)
Active
Application number
CN201711227953.1A
Other languages
Chinese (zh)
Other versions
CN108170664A (en
Inventor
翁永金
李百川
陈第
蔡锐涛
李展铿
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.)
Youmi Technology Co ltd
Original Assignee
Youmi Technology 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 Youmi Technology Co ltd filed Critical Youmi Technology Co ltd
Priority to CN201711227953.1A priority Critical patent/CN108170664B/en
Publication of CN108170664A publication Critical patent/CN108170664A/en
Application granted granted Critical
Publication of CN108170664B publication Critical patent/CN108170664B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities
    • G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
    • 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
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention relates to a key word expansion method and device based on key words. The method comprises the following steps: acquiring first-level keywords covered by the APP to be expanded, and screening out key keywords from the first-level keywords; obtaining a second-stage APP associated with the APP to be expanded according to the APPs searched by the key keywords in the application library platform; further acquiring keywords covered by each second-level APP to obtain a candidate keyword set; then, calculating the comprehensive similarity of each keyword in the candidate keyword set relative to the key keywords; according to the comprehensive similarity and the proportion of each keyword in the candidate keyword set, calculating the similarity score of each keyword in the candidate keyword set; and finally, screening the candidate keyword set based on the similarity score to obtain the associated keywords of the APP to be expanded. The method can automatically develop the keywords related to the APP, thereby realizing mass production and ensuring the development quality.

Description

Key word expansion method and device based on key words
Technical Field
The invention relates to the technical field of data analysis, in particular to a keyword expansion method and device based on key keywords.
Background
With the rapid development of intelligent terminals, the development of the mobile internet software industry is driven. More and more users download various APPs (also called applications) on an application library platform (i.e., an application store) in the smart terminal, and 65% of users search for downloading a desired application through the application store according to a wikipedia data display. Therefore, in order to improve the search quality of the APP developer in the application store, the APP developer needs to make optimization work of the application store. One of the key tasks is to make keyword analysis of the APP to optimize the APP of the user.
At present, based on the specific industry knowledge background of an intelligent terminal application store, keyword expansion of APP is judged and expanded by manpower, and for the manual expansion, expansion quality is greatly influenced by the subjective cognitive level of the manpower, so that the defect that the quality of a keyword expansion result is unstable exists.
Disclosure of Invention
Based on the key word expansion method and device, the key word expansion method and device based on the key words can overcome the defect that the key word expansion quality of the existing application program is unstable.
The scheme provided by the embodiment of the invention comprises the following steps:
a key word expansion method based on key words comprises the following steps:
acquiring first-level keywords covered by the APP to be expanded, and screening out key keywords from the first-level keywords;
obtaining a second-stage APP associated with the APP to be expanded according to the APPs searched by the key keywords in the application library platform; obtaining keywords covered by each second-stage APP, and obtaining a candidate keyword set according to the keywords covered by all the second-stage APPs;
calculating the comprehensive similarity of each keyword in the candidate keyword set relative to the key keywords according to the similarity of each keyword in the candidate keyword set and the corresponding key keywords; acquiring the proportion of each keyword in the candidate keyword set; calculating the similarity score of each keyword in the candidate keyword set according to the proportion and the comprehensive similarity;
screening the candidate keyword set according to the similarity score to obtain associated keywords of the APP to be expanded;
wherein, the keywords covered by APP need to satisfy the conditions: and the search result corresponding to the keyword contains the APP.
A keyword expansion device based on key keywords comprises:
the first word expansion module is used for acquiring first-level keywords covered by the APP to be expanded and screening key keywords from the first-level keywords;
the second word expansion module is used for obtaining a second-level APP related to the APP to be expanded according to the APP information searched by each key keyword on the application library platform; obtaining keywords covered by each second-stage APP, and obtaining a candidate keyword set according to the keywords covered by all the second-stage APPs;
the similarity calculation module is used for calculating the comprehensive similarity of each keyword in the candidate keyword set relative to the key keywords according to the similarity of each keyword in the candidate keyword set and the corresponding key keywords; acquiring the proportion of each keyword in the candidate keyword set; calculating the similarity score of each keyword in the candidate keyword set according to the proportion and the comprehensive similarity;
the keyword screening module is used for screening the candidate keyword set according to the similarity score to obtain associated keywords of the APP to be expanded;
wherein, the keywords covered by APP need to satisfy the conditions: and the search result corresponding to the keyword contains the APP.
A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method as described above.
A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method described above when executing the program.
By implementing the embodiment, after receiving the APP to be expanded, first-level keywords covered by the APP to be expanded are obtained, and key keywords are screened out from the first-level keywords; obtaining a second-stage APP associated with the APP to be expanded according to the APPs searched by the key keywords in the application library platform; further acquiring keywords covered by each second-level APP to obtain a candidate keyword set; then, calculating the comprehensive similarity of each keyword in the candidate keyword set relative to the key keywords; according to the comprehensive similarity and the proportion of each keyword in the candidate keyword set, calculating the similarity score of each keyword in the candidate keyword set; according to the technical scheme, further keyword expansion can be realized based on key keywords of the APP to be expanded according to the APP to be expanded, the keyword expansion breadth is improved, and the keyword expansion quality is guaranteed.
Drawings
FIG. 1 is a schematic flow chart of a keyword expansion method based on key words according to an embodiment;
FIG. 2 is an APP level schematic diagram of a key word expansion method based on key words according to an embodiment;
fig. 3 is a schematic structural diagram of a keyword expansion apparatus based on key words according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The terms "comprises" and "comprising," and any variations thereof, of embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or (module) elements is not limited to only those steps or elements but may alternatively include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
Reference herein to "a plurality" means two or more. "and/or" describes the association relationship of the associated objects, meaning that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments.
Although the steps in the embodiments of the present invention are arranged by using the reference numerals, the order of the steps is not limited to be limited, and the relative order of the steps can be adjusted unless the order of the steps is explicitly described or other steps are required for performing a step.
FIG. 1 is a schematic flow chart of a keyword expansion method based on key words according to an embodiment; as shown in fig. 1, the keyword expansion method based on key words in this embodiment includes the steps of:
s11, obtaining the first-level keywords covered by the APP to be expanded, and screening out the key keywords.
The keywords in the embodiment of the present invention include all keywords that can be used for searching APP on the application library platform, such as chinese characters, english words, or letters, numbers, or other characters, and may also be a combination of several characters. The first-level keywords can be obtained by analyzing historical search information of an application library platform, and the historical search information comprises a mapping relation between the keywords and the APP and can also be pre-specified according to experience values.
Wherein, the keywords covered by APP need to satisfy the conditions: and the search result corresponding to the keyword contains the APP. Namely, each first-level keyword comprises the APP to be expanded in the search result of the application library platform. The key keywords need to satisfy the conditions: in the search results of the key words, the APP to be expanded is ranked first, for example, the APP to be expanded is ranked 10 top.
The key keywords can be obtained by analyzing historical search information of the application library platform, and can also be specified in advance according to experience values.
And S12, obtaining the second-stage APP associated with the APP to be expanded according to the APPs searched by the key keywords in the application library platform. And obtaining keywords covered by each second-level APP, and obtaining a candidate keyword set according to the keywords covered by all the second-level APPs.
Optionally, the step further includes screening the APPs searched by each key keyword on the application library platform, and selecting only the APPs ranked at the top 100 (the number can be specifically set), thereby obtaining the second-stage APP associated with the APP to be expanded. For example, 500 APPs which can be searched by a key keyword on an application library platform are selected, and only the APPs with the top 100 ranks are selected, so that the calculation complexity of subsequent keyword expansion can be reduced, and meanwhile, the later the rank is, the lower the relevance between the keyword and the APPs is, so that the APPs which are ranked later are removed, and the accuracy of keyword expansion can also be ensured.
Wherein, a second level keyword that second level APP covers needs to satisfy the condition: the second-level keywords comprise the second-level APP in the search results of the application library platform.
S13, calculating the comprehensive similarity of each keyword in the candidate keyword set relative to the key keyword according to the key keyword corresponding to each keyword in the candidate keyword set and the similarity of each keyword and the corresponding key keyword; acquiring the proportion of each keyword in the candidate keyword set; and calculating the similarity score of each keyword in the candidate keyword set according to the proportion and the comprehensive similarity.
The similarity of each keyword and each key keyword in the candidate keyword set represents the association degree of the keywords in the same application platform, and can reflect the contact degree of searched APPs, and the similarity of the keywords and the keywords can be predetermined or calculated in real time based on the search records of the application platform. Optionally, the calculating, according to the key words corresponding to the key words in the candidate key word set and the similarity between each key word and the corresponding key word, the comprehensive similarity of each key word in the candidate key word set with respect to the key word includes: the key keywords corresponding to the keywords in the candidate keyword set and the similarity between the keywords and the corresponding key keywords are obtained, and the average value of the similarity between the keywords in the candidate keyword set and the corresponding key keywords is calculated and used as the comprehensive similarity of the keywords in the candidate keyword set relative to the key keywords. The average includes an absolute average and also includes a weighted average.
For example: suppose key words: "shopping", "Taobao";
the keyword "shopping" develops the keywords covered by the APP: [ Beijing, Suningyi purchase ]
The keyword "panning" expands out the keyword that APP covered: [ Jingdong, Tianmao ]
Then the candidate keyword set is [ kyoto, suring easy to purchase, tianmao ]; wherein, the comprehensive similarity of the Beijing east relative to the key keywords in the candidate keyword set is as follows:
sim (kyotong) ═ sim (shopping, kyoton) + sim (naobao, kyoton) ]/2.
The comprehensive similarity of the 'Suningyibui' in the candidate keyword set relative to the key keywords is as follows:
sim (available from suning) ═ sim (shopping, available from suning).
The comprehensive similarity of the "Tianmao" in the candidate keyword set relative to the key keywords is as follows:
sim (cat) is sim (Tanbao, cat).
The proportion of each keyword in the candidate keyword set is determined based on the importance of the keyword to the corresponding APP, and the importance of the keyword to one APP represents ranking information of the APP in the keyword search result. The importance of the keyword to the APP may be obtained in advance through data analysis of historical search record data of the application library platform, or may be a preset importance. If the search result is the former, in an embodiment, the method further includes the step of determining the importance of each keyword to the searched APP in advance according to the historical search record information of the application library platform.
S14, screening the candidate keyword set according to the similarity score to obtain associated keywords of the APP to be expanded.
In one embodiment, a set number of keywords with similarity scores ranked from high to low can be selected from a candidate keyword set to obtain associated keywords of the APP to be expanded; therefore, the associated keywords of the APP to be expanded can be obtained in batches.
In another embodiment, a set number of keyword phrases can be selected from the candidate keyword set according to the sequence of the similarity score from high to low, each keyword phrase includes a plurality of keywords, and the associated keywords of the APP to be expanded are obtained. And a plurality of keyword phrases corresponding to the APP to be expanded can be obtained, so that associated keywords of the APP to be expanded can be conveniently derived in batches.
By the keyword expanding method of the embodiment, after the APP to be expanded is received, the first-level keywords covered by the APP to be expanded are obtained at first, and the key keywords are screened out from the first-level keywords; obtaining a second-stage APP associated with the APP to be expanded according to the APPs searched by the key keywords in the application library platform; further acquiring keywords covered by each second-level APP to obtain a candidate keyword set; then, calculating the comprehensive similarity of each keyword in the candidate keyword set relative to the key keywords; according to the comprehensive similarity and the proportion of each keyword in the candidate keyword set, calculating the similarity score of each keyword in the candidate keyword set; according to the technical scheme, further keyword expansion can be realized based on key keywords of the APP to be expanded according to the APP to be expanded, the keyword expansion breadth is improved, and the keyword expansion quality is guaranteed.
In an embodiment, the process of obtaining the first-level keyword covered by the APP to be expanded may include: acquiring all keywords covered by the APP to be expanded according to the historical search records of the application library platform; and carrying out exception screening on all keywords covered by the APP to be expanded so as to delete the abnormal keywords in the keywords to obtain the first-stage keywords covered by the APP to be expanded. Wherein the abnormal keyword comprises: the search index does not meet at least one of the setting condition, the search result data does not meet the setting condition, the importance does not meet the setting condition, and the number of words does not meet the setting condition.
The search index is obtained by calculating the accumulated times (search amount) of APP search in the application library platform by adopting the keyword within the set statistical time and considering factors such as search magnitude and the like, the search index and the search amount present a forward relation and are estimated approximately empirically, and the search amount corresponding to the search index is as follows:
Figure BDA0001487616140000081
wherein, P is the search index, f (x) represents the non-simple linear growth relationship between the search index and the search quantity.
The search index is less than the set value when the search index does not meet the set condition; the search result data does not meet the set conditions, namely, the number of the APP searched by the keywords is less than the set number; the importance degree is not in accordance with the set condition, namely, APP in the search result of the keyword is ranked relatively later; the word number not meeting the setting condition means that the word number is too short or too long.
Correspondingly, the process of obtaining the keywords covered by each second-level APP may include: acquiring all keywords covered by each second-level APP according to the historical search record of the application library platform; and carrying out exception screening on all keywords covered by each second-level APP to delete the abnormal keywords therein to obtain the keywords covered by the second-level APP.
Further, the specific process of screening out key words from the first-stage key words covered by the APP to be expanded includes: acquiring the importance of each first-stage keyword to the APP to be expanded, and selecting the first-stage keywords with the importance greater than or equal to a first set importance threshold as key keywords covered by the APP to be expanded; the importance of the keywords to the APP to be expanded represents ranking information of the APP to be expanded in the search results of the keywords.
The keyword filtering processing aims at performing abnormal screening on the keywords, for example, the keyword search results are too few, the search index is too low, the search ranking is back, the word number is too short or too long, and the like belong to the abnormal conditions of the keywords, and the abnormal conditions are removed, so that the interference of abnormal data on subsequent expansion is prevented, and the accuracy of keyword expansion is improved.
In one embodiment, the importance of the keyword to the corresponding APP is determined as follows:
according to ranking information of the APP in the keyword search result, the importance of the keyword to the APP is assigned:
V_2(w)＝(15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0.5)
V_3(r)＝(0,1,3,6,10,16,22,30,40,50,65,80,100,120,150,200,∞)
wi＝V_2(w)t；V_3(r)t＜rank≤V_3(r)t+1
wherein i ∈ [1,16 ]](ii) a V _2(w) is an importance weight vector; v _3(r) is a ranking interval vector; infinity represents a positive infinity ranking; rank represents the ranking of APPs in the search results; w is aiRepresenting a keyword kiImportance to APP. For example, APP at keyword kiIs ranked as 2, then the keyword kiThe importance of the APP is wi＝V_2(w)2＝14；V_3(r)2＜rank≤V_3(r)3. Wherein, V _2(w) and V _3(r) can be preset according to different application library platforms.
In an optional embodiment, before obtaining the keywords covered by the APP to be expanded according to the historical search record information of the application library platform, a step of preprocessing the historical search record information of the application library platform is further included. For example, based on the search log information that occurred in the application library platform in the last week, the historical search log information includes keyword information for searching and search result information corresponding to each keyword. Such as keyword search results of the last week, APP information (which may include dimensions of APPID, APP name, affiliated list, etc.), keyword information (which includes dimensions of keyword ID, keyword, search index, search results, etc.).
In an alternative embodiment, the step of preprocessing the historical search record information of the application library platform may comprise:
firstly, acquiring historical search record information of an application library platform in a set time period, and determining a first mapping relation corresponding to each keyword according to the historical search record information; the first mapping relation comprises APP information corresponding to the keyword and ranking information of the APP in the multiple search results of the keyword. Then, according to the first mapping relation of a plurality of keywords in the historical search record information, determining a second mapping relation corresponding to each APP; the second mapping relation comprises keywords corresponding to the APP and the importance of each keyword to the APP, the importance is used for representing ranking information of the APP in the search results of the keywords, and the importance of the keywords to the APP is larger as the APP ranks in the search results of the keywords earlier. Further, a data mapping library corresponding to the application library platform is established according to the first mapping relation and the second mapping relation.
Based on the data mapping library, the obtaining of the first-level keyword corresponding to the APP to be expanded according to the historical search record information of the application library platform may include: and querying the data mapping database, acquiring a second mapping relation corresponding to the APP to be expanded, and acquiring a first-stage keyword corresponding to the APP to be expanded and the importance of the first-stage keyword according to the second mapping relation.
Obtaining the APP searched by each key keyword on the application library platform may include: and querying the data mapping database, acquiring a first mapping relation corresponding to each key word, and obtaining APP information covered by each key word according to the first mapping relation.
In an embodiment, obtaining the second-level APP associated with the APP to be expanded according to the APP searched by each key keyword in the application library platform includes:
obtaining frequency sequencing information of APP in multiple search results corresponding to key words according to multiple search results of the key words in a historical search record within a set historical time period; and acquiring a set number of APPs with the frequency sequence arranged in front as APP information searched by each key keyword. Obtaining an APP matrix according to all key keywords and APP information searched by each key keyword; and counting the occurrence frequency of each APP in the APP matrix, and selecting the APP with the occurrence frequency greater than or equal to a first set frequency in the APP matrix as a second-stage APP associated with the APP to be expanded.
Referring to FIG. 2, the APP to be expanded is a first-stage APP (i.e., APP)(1)) The first-level keyword covered by APP to be expanded is represented as KW(1)And the second-level APP searched by the first-level keyword is represented as the APP(2)The keywords covered by the second-level APP are represented as KW(2)And so on. The key keyword set can be represented as KW(1)′The candidate keyword set can be represented as KW(2)′。
Since the same keyword may be searched for multiple times within a set history period (e.g., within one week), the search result changes with the change of the search time. The search results are counted and summarized to finally obtain a keyword k0Corresponding APP set A (k)0) And a frequency ordering vector V (k)0)，
A(k0)＝(appid1,appid2,…,appidn)
V(k0)＝(count1,count2,…,countn)
Wherein k is0Representing a keyword, countnIndicating the use of a keyword k within a set history period0Searching for appearing appidnCorresponding to the frequency of the app. Wherein, the frequency ranking information of APP in the multiple search results corresponding to the keyword refers to a frequency ranking vector V (k)0) The frequency of said APP.
In an embodiment, after obtaining the second-level APPs associated with the APPs to be expanded, before obtaining the keywords covered by each second-level APP, the method further includes: and acquiring an application list to which the APP to be expanded belongs in the application library platform, and deleting a second-level APP which belongs to a different application list from the APP to be expanded.
In an embodiment, the specific process of obtaining the similarity between each keyword in the candidate keyword set and each key keyword includes:
obtaining a feature vector of each keyword according to the searched APP of each keyword in the candidate keyword set, and obtaining a feature vector of each key keyword according to the searched APP of each key keyword; carrying out One-Hot coding processing on the feature vector of each keyword in the candidate keyword set and the feature vector of each key keyword respectively to obtain a sparse feature vector of the keyword in the candidate keyword set and a sparse feature vector of each key keyword; calculating the similarity between the keywords and each key keyword in the candidate keyword set according to the sparse feature vector of the keywords in the candidate keyword set and the sparse feature vector of each key keyword;
the dimensionality of the sparse feature vector of the key keyword and the dimensionality of the sparse feature vector of each keyword in the candidate keyword set are equal, and the conditions are met: dVM + n is less than or equal to m; m represents the dimension of the feature vector of the key word, n represents the dimension of the feature vector of each key word in the candidate key word set, dVRepresenting dimensions of the sparse feature vector.
For example: the keyword 1 is 'Taobao', the APPs searched by the keyword 'Taobao' are (APP1, APP2 and APP4) and are used as the feature vector of the keyword 1 'Taobao', and the feature vector dimension is 3; the keyword 2 is "shopping", and if the APP retrieved by "shopping" is (APP4, APP5, APP6), the feature vector dimension as the keyword 2 "shopping" is 3; the feature vectors of the two in the real space are (APP1, APP2, APP4, APP5, APP6), the dimensionality is 5 ≦ 3+3, and the sparse feature vectors of the two are obtained as follows: sparse feature vector of keyword 1 "pan bao": (1,1,1,0,0), sparse feature vector of keyword 2 "shopping": (0,0,1,1,1).
Further, an average value of the similarity between the keyword in the candidate keyword set and each key keyword may be calculated as a comprehensive similarity of the keyword in the candidate keyword set with respect to the key keyword.
Based on the above embodiment, optionally, the determination method of the similarity between the ith keyword and the corresponding key keyword in the candidate keyword set is as follows:
Figure BDA0001487616140000121
in the formula, KW(1)′Representing sets of key words, KW(1)′ kRepresenting the Kth key keyword; KW(2)′ iRepresenting the ith keyword in the candidate keyword set; v (KW)(1)′ k)·V(KW(2)′ i) Indicating KW(1)′ kSparse feature vector and KW(2)′ iInner product of sparse feature vectors of (d); i V (KW)(1)′ k)||2||V(KW(2)′ i)||2Indicating KW(1)′ kSparse feature vector and KW(2)′ iIs the product of the 2-norm of the sparse feature vector.
It is understood that the method for calculating the similarity between two keywords includes, but is not limited to, the above algorithm for calculating the similarity based on cosine similarity, and other algorithms for calculating the similarity may also be used.
In an embodiment, obtaining a candidate keyword set according to all keywords covered by the second-level APP includes:
obtaining a keyword matrix according to all keywords covered by the second-stage APP; merging and counting the keywords in the keyword matrix to obtain a candidate keyword set KW(2)′＝(kw(2)′ 1,kw2 (2)′,…,kwn (2)′) And a keyword frequency vector C corresponding to the candidate keyword set(2)＝(c1，c2,…,cn)；
The candidate keyword set KW(2)′The proportion of the ith keyword in the list is as follows:
Figure BDA0001487616140000131
where i ═ 1,2, …, n, denote a candidate keyword set KW(2)′The total number of keywords contained therein.
In an embodiment, calculating a similarity score of each keyword in the candidate keyword set according to the specific gravity and the comprehensive similarity includes: and obtaining the similarity score of the keyword in the candidate keyword set according to the product of the proportion of the keyword in the candidate keyword set and the comprehensive similarity of the keyword relative to the key keyword. Specific examples thereof include: the similarity score of each keyword in the candidate keyword set relative to the key keyword can be calculated by the following formula:
sim(KW(1)′,KW(2)′ i)＝weighti·cosi′
wherein, KW(1)′Representing sets of key words, KW(2)′ iRepresents the ith keyword, weight in the candidate keyword setiRepresenting a set of candidate keywords KW(2)′Specific gravity of the ith keyword, cosi' representing a set of candidate keywords KW(2)′The comprehensive similarity of the ith keyword.
It can be understood that the similarity score of the keyword in the candidate keyword set is obtained according to the product of the specific gravity of the keyword in the candidate keyword set and the comprehensive similarity of the keyword relative to the key keyword, and may be a direct product or a product obtained by multiplying the product by a proportionality coefficient.
And finally, screening the candidate keyword set according to the similarity score to obtain the associated keywords of the APP to be expanded. According to the technical scheme, further keyword expansion can be realized based on the key keywords of the APP to be expanded, the expansion scope of the keywords is improved, and the expansion quality of the keywords is guaranteed. In addition, by the keyword expansion method of the embodiment, the keyword expansion scheme corresponding to the APP to be expanded can be conveniently derived in batches, and the realization efficiency is greatly improved; the mass production is realized, and the expansion quality can be ensured.
The key word expansion method based on key words in the embodiment of the present invention is further described below by taking an apple app store as an example. In the following embodiments, apple app stores are taken as an example, and the other app library platforms have the same principle. The key word expansion method based on the key words comprises the following steps.
1. Keyword content crawling
And acquiring historical search record data of the apple application store in the last week by using the apple developer API, wherein the historical search record data comprises but is not limited to application names, keyword details, keyword search indexes, keyword search results, application lists and the like.
2. Pre-processing of historical keyword search record data
2.1 forward mapping relation between keyword and APP, denoted A (k), representing search result of keyword k, index of index representing actual rank of APP searched by keyword k,
A(k)＝(appid1,appid2,…,appidn)
wherein n is a positive integer.
It should be noted that, in the embodiment of the present invention, the APPs may be identified by appids, and the appids are uniformly allocated by the application library platform and used for identifying different APPs.
2.2, the inverse mapping relationship between app and keywords, denoted as k (a), represents all keywords covered by application a:
K(a)＝(keyword1,...,keywordn)
wherein n is a positive integer.
3. Obtaining key keywords
3.1 noting that the APP to be expanded is APP(1)；
3.2 obtaining APP by K (a)(1)Overlaid keyword set K (APP)(1)) First-level keywords covered by the APP to be expanded;
3.3 keyword set K (APP)(1)) And (5) carrying out exception screening. The data abnormal conditions that the keyword search results are too few, the search index is too low, the search ranking is back, the word number is too short or too long are all the data abnormal conditions, and the data abnormal conditions are eliminated;
3.4 according to A (k), only APP in the search results is selected(1)The key words with the top k ranking are used as key words and recorded as key word set KW(1)′；
4. Keyword expansion keyword
Note KW(1)′ iSet KW of key words(1)′The ith keyword in (1), traverse KW(1)′The method comprises the following steps: 4.1 obtaining keyword KW according to A (k)(1)′ iCorresponding to appid, only take APP at k before ranking, and record as A (KW)(1)′ i)；
4.2 obtaining A (KW) from K (a)(1)′ i) The keywords covered by each APP in the database are marked as K (A (KW)(1)′ i) Merging and counting the keywords to obtain the frequency of the keywords, selecting the first k keywords with the frequency as candidate keywords to obtain a candidate keyword set as follows: KW(2)′＝(kw(2)′ 1,kw2 (2)′,…,kwn (2)′)，
The frequency vector is: c(2)＝(c1,c2,…,cn)；
Defining a set of candidate keywords KW(2)′The specific gravity of the ith keyword is as follows:
Figure BDA0001487616140000151
wherein i is 1,2, …, n, ciAs a set of candidate keywords KW(2)′The frequency of the ith keyword.
4.3 obtaining key words KW according to A (k)(1)′ iWith candidate keyword set KW(2)′And taking the appid corresponding to each keyword as a feature vector of the keyword, and acquiring corresponding sparse feature vectors based on One-Hot coding.
Based on the sparse feature vectors corresponding to the keywords respectively, key keywords KW can be calculated(1)′ kWith candidate keyword set KW(2)′The cosine similarity of the ith keyword is recorded as cosi。
Figure BDA0001487616140000161
In the formula, KW(1)′ kRepresenting the Kth key keyword; KW(2)′ iRepresenting the ith keyword in the candidate keyword set; v (KW)(1)′ k)·V(KW(2)′ i) Indicating KW(1)′ kSparse feature vector and KW(2)′ iInner product of sparse feature vectors of (d); i V (KW)(1)′ k)||2||V(KW(2)′ i)||2Indicating KW(1)′ kSparse feature vector and KW(2)′ iIs the product of the 2-norm of the sparse feature vector.
Calculating the average value of the similarity of the keywords in the candidate keyword set and the corresponding key keywords as the comprehensive similarity of the keywords in the candidate keyword set relative to the key keywords, and recording the comprehensive similarity of the keywords as cosi′。
4.4 calculate similarity scores between keywords as:
sim(KW(1)′,KW(2)′ i)＝weighti·cosi′；
wherein, KW(1)′Representing sets of key words, KW(2)′ iRepresents the ith keyword, weight in the candidate keyword setiRepresenting a set of candidate keywords KW(2)′The proportion of the ith key word in the content,
cosi' representing a set of candidate keywords KW(2)′The comprehensive similarity of the ith keyword.
4.5 finally, set KW of candidate keywords(2)′The medium keywords are subjected to reverse order (from high to low) according to the similarity score, and KW is taken(2)′M in the middle and the top serve as expansion keywords of the APP to be expanded, and therefore the set W of keywords(2)。
In the above steps, 1-2 can be off-line calculation, and are updated periodically, for example, once again every week. And 3-4, performing online calculation, inquiring the data mapping database for each APP name input by the user to obtain a corresponding appid, and further automatically developing the keywords corresponding to the APP in real time.
The technology is applied to APP association expansion of apple stores, and 3 APP expansion effects are tested. Firstly, 20 keywords are manually expanded for each APP, and then the technology is applied to automatically select the first 100 keywords with similarity scores for each APP. The comparison result shows that 80% of the manually selected keywords are automatically selected, and the effectiveness of the technology is proved. Compared with manual expansion, the efficiency of acquiring the keywords by the technology is improved.
Based on the keyword expansion method of the embodiment, after receiving the APP to be expanded, first-level keywords covered by the APP to be expanded are obtained, and key keywords are screened out from the first-level keywords; obtaining a second-stage APP associated with the APP to be expanded according to the APPs searched by the key keywords in the application library platform; further acquiring keywords covered by each second-level APP to obtain a candidate keyword set; then, calculating the comprehensive similarity of each keyword in the candidate keyword set relative to the key keywords; according to the comprehensive similarity and the proportion of each keyword in the candidate keyword set, calculating the similarity score of each keyword in the candidate keyword set; according to the technical scheme, further keyword expansion can be realized based on key keywords of the APP to be expanded according to the APP to be expanded, the keyword expansion breadth is improved, and the keyword expansion quality is guaranteed. In addition, by the keyword expansion method of the embodiment, the keyword expansion scheme corresponding to the APP to be expanded can be conveniently derived in batches, and the realization efficiency is greatly improved; the mass production is realized, and the expansion quality can be ensured.
It should be noted that, for the sake of simplicity, the foregoing method embodiments are described as a series of acts or combinations, but those skilled in the art should understand that the present invention is not limited by the described order of acts, as some steps may be performed in other orders or simultaneously according to the present invention. Further, the above embodiments may be arbitrarily combined to obtain other embodiments.
Based on the same idea as the key word expansion method based on key words in the above embodiment, the present invention further provides a key word expansion device based on key words, which can be used to execute the key word expansion method based on key words. For convenience of description, in the structural schematic diagram of the embodiment of the keyword expansion apparatus based on the key words, only the part related to the embodiment of the present invention is shown, and those skilled in the art will understand that the illustrated structure does not constitute a limitation to the apparatus, and may include more or less components than those illustrated, or combine some components, or arrange different components.
FIG. 3 is a schematic structural diagram of a keyword expansion apparatus based on key keywords according to an embodiment of the present invention; as shown in fig. 3, the keyword expansion apparatus based on key words in this embodiment includes:
the first word expansion module is used for acquiring first-level keywords covered by the APP to be expanded and screening key keywords from the first-level keywords;
the second word expansion module is used for obtaining a second-level APP related to the APP to be expanded according to the APP information searched by each key keyword on the application library platform; obtaining keywords covered by each second-stage APP, and obtaining a candidate keyword set according to the keywords covered by all the second-stage APPs;
the similarity calculation module is used for calculating the comprehensive similarity of each keyword in the candidate keyword set relative to the key keywords according to the similarity of each keyword in the candidate keyword set and the corresponding key keywords; acquiring the proportion of each keyword in the candidate keyword set; calculating the similarity score of each keyword in the candidate keyword set according to the proportion and the comprehensive similarity;
the keyword screening module is used for screening the candidate keyword set according to the similarity score to obtain associated keywords of the APP to be expanded;
wherein, the keywords covered by APP need to satisfy the conditions: and the search result corresponding to the keyword contains the APP.
It should be noted that, in the embodiment of the key word expansion apparatus based on key words in the above example, because the contents of information interaction, execution process, and the like between the modules are based on the same concept as the foregoing method embodiment of the present invention, the technical effect brought by the contents is the same as the foregoing method embodiment of the present invention, and specific contents may refer to the description in the method embodiment of the present invention, and are not described herein again.
In addition, in the embodiment of the key word expansion apparatus based on key words in the above example, the logical division of each program module is only an example, and in practical applications, the above function distribution may be completed by different program modules according to needs, for example, due to the configuration requirements of corresponding hardware or the convenience of implementation of software, that is, the internal structure of the key word expansion apparatus based on key words is divided into different program modules to complete all or part of the above described functions.
It will be understood by those skilled in the art that all or part of the processes of the methods of the above embodiments may be implemented by a computer program, which is stored in a computer readable storage medium and sold or used as a stand-alone product. When executed, the program may perform all or a portion of the steps of the methods of the various embodiments described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
Accordingly, in an embodiment, a storage medium is further provided, on which a computer program is stored, wherein the program, when executed by a processor, implements any of the key expansion methods based on key words as described in the above embodiments.
In addition, the storage medium may be provided in a computer device, and the computer device further includes a processor, and when the processor executes the program in the storage medium, all or part of the steps of the method in the foregoing embodiments can be implemented.
Accordingly, in an embodiment, a computer device is also provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement any one of the key expansion methods based on key words in the above embodiments.
In the above embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments. It is to be understood that the terms "first level," "second level," and the like, as used herein, are used herein to distinguish objects, but the objects are not limited by these terms.
The above-described examples merely represent several embodiments of the present invention and should not be construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims (12)

1. A keyword expansion method based on key words is characterized by comprising the following steps:
acquiring first-level keywords covered by the APP to be expanded, and screening out key keywords from the first-level keywords;
obtaining a second-stage APP associated with the APP to be expanded according to the APPs searched by the key keywords in the application library platform; obtaining keywords covered by each second-stage APP, and obtaining a candidate keyword set according to the keywords covered by all the second-stage APPs;
determining the comprehensive similarity of each keyword in the candidate keyword set relative to the key keywords according to the similarity of each keyword in the candidate keyword set and the corresponding key keywords; the method comprises the following steps: key keywords corresponding to the keywords in the candidate keyword set and the similarity between the keywords and the corresponding key keywords are obtained; the method comprises the following steps of obtaining the similarity between each keyword in a candidate keyword set and a corresponding key keyword, wherein the steps comprise:
obtaining a feature vector of each keyword according to the searched APP of each keyword in the candidate keyword set, and obtaining a feature vector of each key keyword according to the searched APP of each key keyword;
respectively carrying out One-Hot coding processing on the feature vector of each keyword in the candidate keyword set and the feature vector of the corresponding key keyword to obtain a sparse feature vector of the keyword in the candidate keyword set and a sparse feature vector of the corresponding key keyword;
calculating the similarity between the keywords in the candidate keyword set and the corresponding key keywords according to the sparse feature vectors of the keywords in the candidate keyword set and the sparse feature vectors of the corresponding key keywords;
acquiring the proportion of each keyword in the candidate keyword set according to the importance of each keyword in the second-stage APP corresponding to the keyword in the candidate keyword set; calculating the similarity score of each keyword in the candidate keyword set according to the proportion and the comprehensive similarity;
screening the candidate keyword set according to the similarity score to obtain associated keywords of the APP to be expanded;
wherein, the keywords covered by APP need to satisfy the conditions: and the search result corresponding to the keyword contains the APP.
2. The keyword expansion method based on key words according to claim 1, wherein the screening of key words comprises:
acquiring the importance of each first-stage keyword to the APP to be expanded, and selecting the first-stage keywords with the importance greater than or equal to a first set importance threshold as key keywords covered by the APP to be expanded;
the importance of the keywords to the APP to be expanded represents ranking information of the APP to be expanded in the search results of the keywords.
3. The key word expansion method based on key words of claim 2, wherein the obtaining of the second-level APP associated with the APP to be expanded according to the APP searched by each key word on the application library platform comprises:
obtaining frequency sequencing information of APP in multiple search results corresponding to key words according to multiple search results of the key words in a historical search record within a set historical time period; acquiring a set number of APPs with a frequency sequence arranged in front as the APPs searched by the key keywords;
obtaining an APP matrix according to all key keywords and APPs searched by each key keyword; and counting the occurrence frequency of each APP in the APP matrix, and selecting the APP with the occurrence frequency greater than or equal to the set frequency in the APP matrix as a second-stage APP associated with the APP to be expanded.
4. The keyword expansion method based on key words of claim 3, wherein after obtaining the second-level APP associated with the APP to be expanded, before obtaining the keywords covered by each second-level APP, the method further comprises:
and acquiring an application list to which the APP to be expanded belongs in the application library platform, and deleting a second-level APP which belongs to a different application list from the APP to be expanded.
5. The keyword expansion method based on key words according to claim 1,
determining the comprehensive similarity of each keyword in the candidate keyword set relative to the key keywords according to the similarity of each keyword in the candidate keyword set and the corresponding key keywords, and further comprising:
and calculating the average value of the similarity of each keyword in the candidate keyword set and the corresponding key keyword as the comprehensive similarity of each keyword in the candidate keyword set relative to the key keyword.
6. The keyword expansion method based on key words according to claim 1,
calculating the similarity between the ith keyword and the corresponding key keyword in the candidate keyword set by the following formula:
Figure FDA0002946589680000031
in the formula, KW(1)′ kRepresenting the Kth key keyword; KW(2)′ iRepresenting the ith keyword in the candidate keyword set; v (KW)(1)′ k)·V(KW(2)′ i) Indicating KW(1)′ kSparse feature vector and KW(2)′ iInner product of sparse feature vectors of (d); i V (KW)(1)′ k)||2||V(KW(2)′ i)||2Indicating KW(1)′ kSparse feature vector and KW(2)′ iIs the product of the 2-norm of the sparse feature vector.
7. The key word expansion method based on key words of claim 5, wherein obtaining a candidate key word set according to all the key words covered by the second-level APP comprises:
obtaining a keyword matrix according to all keywords covered by the second-stage APP;
merging and counting the keywords in the keyword matrix to obtain a candidate keyword set KW(2)′＝(kw(2)′ 1,kw(2)′2,…,kw(2)′ n) And a keyword frequency vector C corresponding to the candidate keyword set(2)＝(c1,c2,…,cn) (ii) a Each element of the keyword frequency vector corresponds to the occurrence frequency of each keyword in the candidate keyword set respectively;
the candidate keyword set KW(2)′The proportion of the ith key word is as follows:
Figure FDA0002946589680000032
where i ═ 1,2, …, n, denote a candidate keyword set KW(2)′The total number of keywords contained therein.
8. The key word expansion method based on key words according to any one of claims 1 to 7, wherein calculating the similarity score of each key word in the candidate key word set according to the specific gravity and the comprehensive similarity comprises:
and obtaining the similarity score of the keyword in the candidate keyword set according to the product of the proportion of the keyword in the candidate keyword set and the comprehensive similarity of the keyword relative to the key keyword.
9. The key word expansion method based on key words according to any one of claims 1 to 7, wherein the step of screening the candidate key word set according to the similarity score to obtain associated key words of APP to be expanded comprises the steps of:
selecting a set number of keywords with the similarity scores ranked from high to low from the candidate keyword set to obtain associated keywords of the APP to be expanded;
alternatively, the first and second electrodes may be,
and selecting a set number of keyword phrases from the candidate keyword set according to the sequence of the similarity scores from high to low, wherein each keyword phrase comprises a plurality of keywords, and obtaining associated keywords of the APP to be expanded.
10. A keyword expansion device based on key keywords is characterized by comprising:
the first word expansion module is used for acquiring first-level keywords covered by the APP to be expanded and screening key keywords from the first-level keywords;
the second word expansion module is used for obtaining a second-level APP related to the APP to be expanded according to the APP information searched by each key keyword on the application library platform; obtaining keywords covered by each second-stage APP, and obtaining a candidate keyword set according to the keywords covered by all the second-stage APPs;
the similarity calculation module is used for calculating the comprehensive similarity of each keyword in the candidate keyword set relative to the key keywords according to the similarity of each keyword in the candidate keyword set and the corresponding key keywords; the module is specifically configured to: key keywords corresponding to the keywords in the candidate keyword set and the similarity between the keywords and the corresponding key keywords are obtained; the method comprises the following steps of obtaining the similarity between each keyword in a candidate keyword set and a corresponding key keyword, wherein the steps comprise:
obtaining a feature vector of each keyword according to the searched APP of each keyword in the candidate keyword set, and obtaining a feature vector of each key keyword according to the searched APP of each key keyword;
respectively carrying out One-Hot coding processing on the feature vector of each keyword in the candidate keyword set and the feature vector of the corresponding key keyword to obtain a sparse feature vector of the keyword in the candidate keyword set and a sparse feature vector of the corresponding key keyword;
calculating the similarity between the keywords in the candidate keyword set and the corresponding key keywords according to the sparse feature vectors of the keywords in the candidate keyword set and the sparse feature vectors of the corresponding key keywords; acquiring the proportion of each keyword in the candidate keyword set according to the importance of each keyword in the second-stage APP corresponding to the keyword in the candidate keyword set; calculating the similarity score of each keyword in the candidate keyword set according to the proportion and the comprehensive similarity;
the keyword screening module is used for screening the candidate keyword set according to the similarity score to obtain associated keywords of the APP to be expanded;
wherein, the keywords covered by APP need to satisfy the conditions: and the search result corresponding to the keyword contains the APP.
11. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 9.
12. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the steps of the method of any one of claims 1 to 9 are implemented when the program is executed by the processor.
CN201711227953.1A 2017-11-29 2017-11-29 Key word expansion method and device based on key words Active CN108170664B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711227953.1A CN108170664B (en) 2017-11-29 2017-11-29 Key word expansion method and device based on key words

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711227953.1A CN108170664B (en) 2017-11-29 2017-11-29 Key word expansion method and device based on key words

Publications (2)

Publication Number Publication Date
CN108170664A CN108170664A (en) 2018-06-15
CN108170664B true CN108170664B (en) 2021-04-09

Family

ID=62524152

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711227953.1A Active CN108170664B (en) 2017-11-29 2017-11-29 Key word expansion method and device based on key words

Country Status (1)

Country Link
CN (1) CN108170664B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113128209B (en) * 2021-04-22 2023-11-24 百度在线网络技术（北京）有限公司 Method and device for generating word stock
CN115630154B (en) * 2022-12-19 2023-05-05 竞速信息技术(廊坊)有限公司 Big data environment-oriented dynamic abstract information construction method and system

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103473104A (en) * 2013-09-24 2013-12-25 北京大学 Method for discriminating re-package of application based on keyword context frequency matrix
CN104216995A (en) * 2014-09-10 2014-12-17 北京金山安全软件有限公司 Information processing method and device
CN105302810A (en) * 2014-06-12 2016-02-03 北京搜狗科技发展有限公司 Information search method and apparatus
CN106649818A (en) * 2016-12-29 2017-05-10 北京奇虎科技有限公司 Recognition method and device for application search intentions and application search method and server
CN106649323A (en) * 2015-10-29 2017-05-10 北京国双科技有限公司 Method and device for recommending keyword

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101295319B (en) * 2008-06-24 2010-06-02 北京搜狗科技发展有限公司 Method and device for expanding query, search engine system
CN102110174B (en) * 2011-04-11 2013-04-03 重庆大学 Keyword-based WEB server expansion search method
US9280595B2 (en) * 2012-08-30 2016-03-08 Apple Inc. Application query conversion
CN103902720B (en) * 2014-04-10 2017-11-21 北京博雅立方科技有限公司 The expansion word acquisition methods and device of a kind of keyword
CN104199982B (en) * 2014-09-25 2017-09-26 北京金山安全软件有限公司 Method and device for displaying search keywords
CN104573009B (en) * 2015-01-08 2018-08-24 南通大学 A kind of method of domain knowledge base attribute extension
CN105677695B (en) * 2015-09-28 2019-03-08 杭州圆橙科技有限公司 A method of the calculating mobile application similitude based on content
US10331680B2 (en) * 2015-12-28 2019-06-25 Samsung Electronics Co., Ltd. Ranking of search results

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103473104A (en) * 2013-09-24 2013-12-25 北京大学 Method for discriminating re-package of application based on keyword context frequency matrix
CN105302810A (en) * 2014-06-12 2016-02-03 北京搜狗科技发展有限公司 Information search method and apparatus
CN104216995A (en) * 2014-09-10 2014-12-17 北京金山安全软件有限公司 Information processing method and device
CN106649323A (en) * 2015-10-29 2017-05-10 北京国双科技有限公司 Method and device for recommending keyword
CN106649818A (en) * 2016-12-29 2017-05-10 北京奇虎科技有限公司 Recognition method and device for application search intentions and application search method and server

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
查询扩展技术进展与展望;黄名选等;《计算机应用与软件》;20071115;第24卷(第11期);1-4,8 *

Also Published As

Publication number Publication date
CN108170664A (en) 2018-06-15

Similar Documents

Publication Publication Date Title
CN109086394B (en) Search ranking method and device, computer equipment and storage medium
CN106951571B (en) Method and device for labeling application with label
CN109766492B (en) Learning recommendation method, device, equipment and readable medium
US10346496B2 (en) Information category obtaining method and apparatus
CN108182200B (en) Keyword expansion method and device based on semantic similarity
CN110765348B (en) Hot word recommendation method and device, electronic equipment and storage medium
JP2020170538A (en) Method, apparatus and program for processing search data
CN113268641B (en) User data processing method based on big data and big data server
JP2005322165A (en) Retrieval keyword presentation method, device, and program
CN111859004A (en) Retrieval image acquisition method, device, equipment and readable storage medium
CN114223012A (en) Push object determination method and device, terminal equipment and storage medium
CN111241381A (en) Information recommendation method and device, electronic equipment and computer-readable storage medium
CN108170664B (en) Key word expansion method and device based on key words
CN115238173A (en) Behavior analysis and medical service pushing method, equipment and medium based on big data
CN110008396B (en) Object information pushing method, device, equipment and computer readable storage medium
CN111160699A (en) Expert recommendation method and system
CN110968802A (en) User characteristic analysis method, analysis device and readable storage medium
CN113220557A (en) Cold-start recommendation model evaluation method, system, computer device and storage medium
JP4891638B2 (en) How to classify target data into categories
CN108170665B (en) Keyword expansion method and device based on comprehensive similarity
CN110019783B (en) Attribute word clustering method and device
CN111382342B (en) Method, device and equipment for acquiring hot search words and storage medium
CN112579912A (en) Searching method, electronic equipment and computer storage medium
CN114610859A (en) Product recommendation method, device and equipment based on content and collaborative filtering
CN114218259A (en) Multi-dimensional scientific information search method and system based on big data SaaS

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