CN108021640A - Keyword expanding method and device based on associated application - Google Patents

Keyword expanding method and device based on associated application Download PDF

Info

Publication number
CN108021640A
CN108021640A CN201711227933.4A CN201711227933A CN108021640A CN 108021640 A CN108021640 A CN 108021640A CN 201711227933 A CN201711227933 A CN 201711227933A CN 108021640 A CN108021640 A CN 108021640A
Authority
CN
China
Prior art keywords
app
stage
keywords
keyword
level
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201711227933.4A
Other languages
Chinese (zh)
Other versions
CN108021640B (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.)
Umi-Tech Co Ltd
Original Assignee
Umi-Tech 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 Umi-Tech Co Ltd filed Critical Umi-Tech Co Ltd
Priority to CN201711227933.4A priority Critical patent/CN108021640B/en
Publication of CN108021640A publication Critical patent/CN108021640A/en
Application granted granted Critical
Publication of CN108021640B publication Critical patent/CN108021640B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/953Querying, e.g. by the use of web search engines
    • G06F16/9535Search customisation based on user profiles and personalisation

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 present invention relates to keyword expanding method and device based on associated application.The described method includes:Obtain the first order keyword of APP to be expanded covering, according to each first order keyword search to APP obtain associated second level APP;Obtain the second level keyword of each second level APP covering, according to each second level keyword search to APP obtain the associated third level APP of APP to be expanded;One candidate key set of words is obtained according to each third level APP keywords covered;Determine similarities of each acquisition third level APP relative to second level APP, obtain the proportion shared by each keyword in candidate key set of words;Calculate the similarity score of each keyword in candidate key set of words;Candidate key set of words is screened according to the similarity score, obtains the association keyword of APP to be expanded.The present invention can expand out the relevant keywords of APP automatically, both realize volume production, while ensure that expansion quality.

Description

Keyword expansion method and device based on associated application
Technical Field
The invention relates to the technical field of data analysis, in particular to a keyword expansion method and device based on associated application.
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 (applications, also called applications) in 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 the 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 method and the device, the keyword expansion method and the device based on the associated application can overcome the defect that the keyword expansion quality of the existing application program is unstable.
The scheme provided by the embodiment of the invention comprises the following steps:
a keyword expansion method based on associated application comprises the following steps:
acquiring first-stage keywords covered by the APP to be expanded, and acquiring second-stage APPs related to the APPs to be expanded according to the APPs searched by the first-stage keywords in the application library platform;
acquiring second-stage keywords covered by each second-stage APP, and obtaining third-stage APPs related to the APPs to be expanded according to the APPs searched by each second-stage keyword in the application library platform; obtaining keywords covered by each third-stage APP, and obtaining a candidate keyword set according to the keywords covered by each third-stage APP;
determining the similarity of each third-stage APP relative to the second-stage APP, and 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 similarity and the proportion;
screening the candidate keyword set according to the similarity score to obtain associated keywords of the APP to be expanded;
wherein, the keyword that APP covered needs to satisfy the condition: and the search result corresponding to the keyword contains the APP.
A keyword expanding device based on associated application comprises:
the application expansion module is used for acquiring first-stage keywords covered by the APP to be expanded, and obtaining second-stage APPs related to the APPs to be expanded according to the APPs searched by the first-stage keywords in the application library platform;
the candidate word expansion module is used for acquiring second-stage keywords covered by each second-stage APP, and obtaining third-stage APPs related to the APPs to be expanded according to the APPs searched by the second-stage keywords in the application library platform; obtaining keywords covered by each third-stage APP, and obtaining a candidate keyword set according to the keywords covered by each third-stage APP;
the similarity calculation module is used for determining the similarity of each third-stage APP relative to the second-stage APP and 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 similarity and the proportion;
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 the APP to be expanded is received, the first-stage keywords covered by the APP to be expanded can be determined, and then the second-stage APP (the APP of the competitive products) is obtained according to the APP information corresponding to each first-stage keyword; further, a third-level APP related to the APP to be expanded can be obtained through the second-level keywords covered by the second-level APPs and the APP information corresponding to the second-level keywords; obtaining a candidate keyword set according to the keywords covered by each third-stage APP; determining the similarity of each third-stage APP relative to the second-stage APP, and calculating the similarity score of each keyword in the candidate keyword set by combining the proportion of each keyword in the candidate keyword set; according to the technical scheme, expansion of the keywords can be achieved based on the APP to be expanded, and the quality of keyword expansion can be improved. 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.
Drawings
FIG. 1 is a schematic flow chart diagram of a keyword expansion method based on a related application according to an embodiment;
FIG. 2 is an APP level schematic diagram of a keyword expansion method based on associated applications according to an embodiment;
FIG. 3 is a schematic structural diagram of a keyword expansion apparatus based on a related application 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 diagram of a keyword expansion method based on a related application according to an embodiment; as shown in fig. 1, the keyword expansion method based on the associated application in this embodiment includes the steps of:
s11, obtaining first-stage keywords covered by the APP to be expanded, and obtaining second-stage APPs related to the APPs to be expanded according to the APPs searched by the first-stage keywords in the application library platform.
The keywords in the embodiment of the present invention include all characters that can be used for searching for 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 mapping relations between the keywords and the APP and can also be pre-specified according to empirical 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.
S12, obtaining second-stage keywords covered by each second-stage APP, and obtaining third-stage APPs related to the APPs to be expanded according to the APPs searched by the second-stage keywords in the application library platform. And obtaining keywords covered by each third-stage APP, and obtaining a candidate keyword set according to the keywords covered by each third-stage APP.
The second-level keywords may be obtained by analyzing historical search information of the application library platform, or may be pre-specified according to empirical values. A second level keyword covered by a second level APP, which needs to satisfy the condition: the second-level keywords comprise the second-level APP in the search results of the application library platform.
The keywords covered by the third-level APP can be obtained by analyzing historical search information of an application library platform, and can also be pre-specified according to empirical values. A keyword covered by a third-level APP, which needs to satisfy the condition: and the keyword contains the third-level APP in the search result of the application library platform.
S13, determining the similarity of each third-stage APP relative to the second-stage APP, and 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 similarity and the proportion.
And the similarity of the third-stage APP relative to the second-stage APP represents the comprehensive association degree of the third-stage APP and the corresponding second-stage APP. In an embodiment, if one second-stage APP corresponding to a third-stage APP is available, obtaining a similarity between the third-stage APP and the corresponding second-stage APP, as a similarity of the third-stage APP with respect to the second-stage APP; if the number of the second-stage APPs corresponding to the third-stage APP is more than two, the similarity between the third-stage APP and each corresponding second-stage APP is respectively obtained, so that a similarity mean value is calculated, and the similarity mean value is used as the similarity of the third-stage APP relative to the second-stage APPs. The similarity between the third-stage APP and the single second-stage APP may be predetermined, or may be calculated in real time based on a search record of an application platform. The calculating of the similarity mean value comprises calculating an absolute mean value and calculating a weighted mean value.
The proportion of each keyword in the candidate keyword set is determined based on the importance of the keyword to the third-level APP, and the importance of the keyword to one APP represents ranking information of the APP in a search result of the keyword. 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 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.
And S14, screening the candidate keyword set according to the similarity score to obtain the 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 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-stage keywords covered by the APP to be expanded can be determined, and then the second-stage APP (the competitive product APP) is obtained according to the APP information corresponding to each first-stage keyword; further, a third-level APP related to the APP to be expanded can be obtained through the second-level keywords covered by the second-level APPs and the APP information corresponding to the second-level keywords; obtaining a candidate keyword set according to the keywords covered by each third-stage APP; determining the similarity of each third-stage APP relative to the second-stage APP, and calculating the similarity score of each keyword in the candidate keyword set by combining the proportion of each candidate keyword; according to the technical scheme, expansion of the keywords can be achieved based on the APP to be expanded, and the quality of keyword expansion can be improved.
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 is abnormal, the data of the keyword search result is abnormal, the APP ranks in the search result to be abnormal, and the keyword is at least one characteristic of the character number abnormality.
The search index is obtained by calculating the accumulated times (search volume) of APP search in the application library platform by adopting the keyword within the set statistical time and considering the factors such as search magnitude, the search index and the search volume present a positive relationship and are estimated approximately from experience, and the search volume corresponding to the search index is as follows:
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 abnormal, namely the search index is smaller than a set value; the abnormal search result means that the number of the APPs searched by the keywords is less than the set number; the importance degree abnormality means that the APP ranks relatively later in the search results of the keywords; word count anomalies mean that the number of words is too short or too long.
Correspondingly, the process of obtaining the second-level 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 second-level keywords covered by the second-level APP.
The process of obtaining the keywords covered by each third-level APP may include: acquiring all keywords covered by each third-stage APP according to the historical search record of the application library platform; and carrying out exception screening on all keywords covered by each third-stage APP to delete the abnormal keywords therein to obtain the keywords covered by the third-stage APP.
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 an embodiment, the method further includes a step of predetermining importance of the keyword to the corresponding APP, and specifically includes: 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,∞)
w i =V_2(w) t ;V_3(r) t <rank≤V_3(r) t+1
wherein i is E [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 a i Representing a keyword k i Importance to APP. For example, APP at keyword k i Is ranked as 2, then the keyword k i The importance of the APP is w i =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.
The obtaining of APP information covered by each first-level key keyword on the application library platform may include: and querying the data mapping database, acquiring a first mapping relation corresponding to each first-level key word, and obtaining APP information covered by each first-level 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 first-level keyword in the application library platform includes:
obtaining the APP frequency sequencing information in the multiple search results corresponding to the first-stage keywords according to the multiple search results of the first-stage keywords in the historical search records within a set historical time period; and acquiring a set number of APPs with the frequency sequence arranged at the front as APP information searched by each first-level keyword. Obtaining an APP matrix according to all the first-level keywords and APP information searched by each first-level 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.
In an embodiment, the obtaining a third-level APP associated with an APP to be expanded according to the APP searched by each second-level keyword in the application library platform includes:
obtaining the APP frequency sequencing information in the multiple search results corresponding to the second-level keywords according to the multiple search results of the second-level keywords in the historical search records within the set historical time period; acquiring a set number of APPs with a frequency sequence arranged in front as APP information searched by each second-level keyword; obtaining an APP matrix according to all the second-level keywords and APP information searched by all the second-level keywords; 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 second set frequency in the APP matrix as a third-stage APP associated with the second-stage APP.
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 k 0 Corresponding APP set A (k) 0 ) And a frequency ordering vector V (k) 0 ),
A(k 0 )=(appid 1 ,appid 2 ,…,appid n )
V(k 0 )=(count 1 ,count 2 ,…,count n )
Wherein k is 0 Representing a keyword, count n Indicating the use of a keyword k within a set history period 0 Searching for appearing appid n Corresponding 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 corresponding to the APP in (1).
In an embodiment, after obtaining the second-level APPs, before obtaining the second-level 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.
Optionally, after obtaining the third-level APPs, before obtaining the keywords covered by each third-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 third-stage APP which belongs to an application list different from the APP to be expanded. Therefore, the accuracy of the associated APP of the APP can be improved, and the accuracy of subsequent keyword expansion is improved.
In an embodiment, the similarity between each third-stage APP and a single second-stage APP is a similarity calculated in real time, and the specific calculation process includes:
obtaining a feature vector of the second-stage APP according to the second-stage keywords covered by the second-stage APP, and obtaining a feature vector of each third-stage APP according to the keywords covered by each third-stage APP; processing the feature vector of the second-stage APP and the feature vector of the third-stage APP through One-Hot coding to obtain a sparse feature vector of the second-stage APP and a sparse feature vector of the third-stage APP; and calculating the similarity between each third-stage APP and the corresponding second-stage APP according to the sparse feature vector of the second-stage APP and the sparse feature vector of the third-stage APP. Wherein, the sparse eigenvector of the second-level APP is equal to the sparse eigenvector of the third-level APP in dimensionality, and satisfies the condition: d V M + n is less than or equal to m; m denotes the dimension of the feature vector of the second stage APP, n denotes the dimension of the feature vector of the third stage APP, d V Representing dimensions of the sparse feature vector.
For example: for example, the APP to be expanded is the APP (1) Suppose its corresponding second stage APP comprises (APP) (2) 1 、APP (2) 2 ) Wherein the second stage of APPAPP (2) 1 The covered key word is (KW) (2) 1 ,KW (2) 2 ,KW (2) 3 ) Using the obtained product as a second-stage APPAPP (2) 1 The feature vector dimension of (3); second stage of APPAPP (2) 2 The covered key word is (KW) (2) 2 ,KW (2) 3 ,KW (2) 4 ,KW (2) 5 ) Using the obtained product as a second-stage APPAPP (2) 2 The feature vector dimension of (2) is 4.
Further, a second stage of APPAPP (2) 1 The corresponding third-stage APP comprises (APP) (3) 1 ,APP (3) 2 ,APP (3) 3 ) (ii) a Second stage of APPAPP (2) 2 The corresponding third-stage APP comprises (APP) (3) 3 ,APP (3) 4 ,APP (3) 5 ) (ii) a From this, a third-level APP set (APP) is obtained (3) 1 ,APP (3) 2 ,APP (3) 3 ,APP (3) 4 ,APP (3) 5 ). In the third-level APP set, the APPs (3) 1 Corresponding second stage APP Only APP (2) 1 Thus, APP (3) 1 Similarity to second-stage APP, i.e. APP (3) 1 With APP (2) 1 The similarity of (2); APP (3) 3 Corresponding second-stage APP has APP (2) 1 And APP (2) 2 Thus, obtaining APP separately (3) 3 With APP (2) 1 Similarity of (A) and (B) APP (3) 3 With APP (2) 2 Calculating a similarity mean value by using the similarity mean value, and taking the similarity mean value as the APP (3) 3 Similarity to second-stage APP.
Further, third stage APPAPP (3) 1 The covered key word is (KW) (3) 1 ,KW (3) 2 ,KW (3) 3 ) This is used as the third-stage APPAPP (3) 1 The feature vector dimension of (3); third stage APPAPP (3) 2 The covered key word is (KW) (3) 4 ,KW (3) 2 ,KW (3) 3 ,KW (3) 5 ) This is used as the third-stage APPAPP (3) 2 The feature vector dimension of (2) is 4. Wherein, KW (3) 2 And KW (2) 2 Are the same keyword.
Thus, the second stage of APPAPP (2) 1 Characteristic vector (KW) (2) 1 ,KW (2) 2 ,KW (2) 3 ) Third grade of APPAP (3) 1 Characteristic vector (KW) (3) 1 ,KW (3) 2 ,KW (3) 3 ),KW (3) 2 And KW (2) 2 The feature vectors are the same keyword, so that the feature vectors formed by the two in the real number space are (KW) (2) 1 ,KW (2) 2 ,KW (2) 3 ,KW (3) 1 ,KW (3) 3 ) Dimension of 5&=3+3, and the sparse feature vectors obtained by the two methods are respectively: second stage of APPAPP (2) 1 Sparse feature vector of (2): (1,1,1,0,0), third stage APPPP (3) 1 Sparse feature vector of (2): (0,1,0,1,1).
Based on the above embodiment, optionally, the similarity between each third-stage APP and a single second-stage APP is calculated by the following formula:
in the formula, APP (2) t Represents the t-th second-stage APP; s (3) i Represents the ith third-stage APP; v (APP) (2) t )·V(S (3) i ) Denotes APP (2) t Sparse feature vector of (1) and (S) (3) i Inner product of sparse feature vectors of (d); i V (APP) (2) t )|| 2 ||V(S (3) i )|| 2 Denotes APP (2) t Sparse feature vector of (1) and (S) (3) i Is the product of the 2-norm of the sparse feature vector.
It is understood that the calculation method of the similarity between two APPs includes, but is not limited to, the above algorithm for calculating the similarity based on the cosine similarity, and other algorithms for calculating the similarity may also be used.
In an embodiment, obtaining a candidate keyword set according to the keywords covered by each third-level APP includes: and obtaining a keyword matrix associated with the second-stage APP according to the third-stage APP associated with the second-stage APP and the keywords covered by each third-stage APP. Merging and counting the keywords in the keyword matrix to obtain candidate relationsSet of key words KW (3) =(kw (3) 1 ,kw (3) 2 ,…,kw (3) n ) And the corresponding keyword frequency vector is C (3) =(c 1 ,c 2 ,…,c n )。
Further, the proportion of each keyword in the candidate keyword set may be: determining a set of candidate keywords KW (3) The proportion of the ith keyword in the list is as follows:
in the formula, i =1,2, …, n, n represents a candidate keyword set KW (3) The total number of keywords contained therein.
In an embodiment, the calculating a similarity score of each keyword in the candidate keyword set according to the similarity and the specific gravity includes: and obtaining the similarity score of the keywords in the candidate keyword set according to the product of the proportion of the keywords in the candidate keyword set and the similarity of the third-stage APP corresponding to the keywords relative to the second-stage APP. Specific examples thereof include: calculating the similarity score of each keyword in the candidate keyword set by the following formula:
score(kw (3) i )=V (1) i V (2) i
wherein kw is (3) i Representing a set of candidate keywords KW (3) The ith keyword in (1), V (1) i Represents kw (3) i Similarity, V, of the corresponding third-stage APP relative to the second-stage APP (2) i Represents kw (3) i The occupied specific gravity; i =1,2, …, n, n represents a candidate keyword set KW (3) The total number of keywords contained therein.
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 similarity of the third-stage APP corresponding to the keyword relative to the second-stage APP, that is, the similarity score may be a direct product or a product obtained by multiplying the similarity score 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. Above-mentioned technical scheme can realize the extension of keyword based on the article APP that competes according to treating the APP that extends, can improve the quality of keyword extension. 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 method for expanding keywords based on associated applications according to 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 keyword expansion method based on the associated application 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 of keywords to APP, denoted A (k), represents the search results for keyword k, the index of index for appid represents the actual ranking of searching APP with keyword k,
A(k)=(appid 1 ,appid 2 ,…,appid n ) (1-1)
wherein n is a positive integer.
It should be noted that, in the embodiment of the present invention, the APP may be identified by an appid, and the appid is uniformly allocated by the application library platform and is used to identify different APPs.
2.2 The inverse mapping relationship between app and keyword, denoted as K (a), represents all keywords covered by application a:
K(a)=(keyword 1 ,...,keyword n ) (1-2)
wherein n is a positive integer.
3. Obtaining contest APP (i.e. associated second level APP)
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 obtaining keyword set K (APP) by A (K) (1) ) Each keyword in the set corresponds to appid, and is marked as A (K (APP) (1) ));
3.5 pairs of A (K (APP) (1) ) ) performing merging statistics, and taking the appid n before the frequency ranking as the APP set S (1)′
3.6 reject APP set S (1)′ Neutralization of APP (1) The APP which do not belong to the same application list only takes k APP as the competitive product APP and records as the competitive product APP set S (1) I.e. the second level APP associated with the APP to be extended.
APP expansion keywords
Note S (1) i Aggregating APP for an item of contest S (1) The ith APP in the game is traversed to the competitive product APP set S (1) The method comprises the following steps:
4.1 obtaining Association appid
Obtaining the competitive products in the first 5 steps in the step 3Is marked as a third-level APP, and the corresponding set uses S (3) Represents:
A(K(S (1) i ))=(appid 1 ,…,appid n ) (4-1)
further, the keyword matrix covered by (4-1) can be obtained:
4.2 extracting the feature vector.
Will compete for goodsCharacteristic vector of (1) and S (3) The feature vectors covered by each APP (namely, the feature vectors and the corresponding row of keywords in (4-2)) are subjected to One-Hot coding, and thus the competitive products are obtainedSparse feature vector V (S) (1) i ) And S (3) Sparse feature vector V (S) covered by each APP (3) i )。
For example: presence vector a = (kw) 1 ,kw 2 ,kw 3 ) And B = (kw) 3 ,kw 4 ,kw 5 ) If both are subjected to One-Hot encoding, then the eigenvectors f = (kw) in the real number space R of both 1 ,kw 2 ,kw 3 ,kw 4 ,kw 5 ) Are a '= (1,1,1,0,0) and B' = (0,0,1,1,1), respectively.
4.3 calculate APP similarity.
Based on the 4.2 results, S was calculated (3) Middle APP and S (1) i The similarity of (c) is as follows:
in the formula, S (3) j Denotes S (3) The jth third stage APP in (1); v (S) (3) j )·V(S (1) i ) Denotes S (3) j Sparse feature vector of (1) and (S) (1) i Inner product of sparse feature vectors of (d); i V (S) (3) j )|| 2 ||V(S (1) i )|| 2 To representS (3) j Sparse feature vector of (1) and (S) (1) i Is the product of the 2-norm of the sparse feature vector.
Merging and counting the keywords in (4-2) to obtain a candidate keyword set KW (3) =(kw (3) 1 ,kw (3) 2 ,…,kw (3) n ) And the corresponding frequency vector is C (3) =(c 1 ,c 2 ,…,c n );
The candidate keyword set KW (3) The specific gravity of the ith keyword is as follows:
where i =1,2, …, n, n represents a candidate keyword set KW (3) The total number of keywords contained therein.
4.4 calculating the similarity of the candidate keywords relative to the APP.
According to the similarity of (4-3) and the specific gravity of (4-4), a candidate keyword set KW can be calculated (3) And the similarity score of each keyword relative to the APP to be expanded.
Finally, for candidate keyword set KW (3) The medium keywords are subjected to reverse order (from high to low) according to the similarity score, and KW is taken (3) M in the middle and the top serve as expansion keywords of the APP to be expanded, and therefore the set W of keywords (1)
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 keyword 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.
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 related application-based keyword expansion method in the above embodiment, the present invention further provides a related application-based keyword expansion apparatus, which can be used to execute the related application-based keyword expansion method. For convenience of description, in the structural schematic diagram of the embodiment of the device expanding based on the keywords of the associated application, 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 device, 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 related applications according to an embodiment of the present invention; as shown in fig. 3, the keyword expansion apparatus based on the associated application of this embodiment includes:
the application expansion module is used for acquiring first-stage keywords covered by the APP to be expanded, and obtaining second-stage APPs related to the APPs to be expanded according to the APPs searched by the first-stage keywords in the application library platform;
the candidate word expansion module is used for acquiring second-stage keywords covered by each second-stage APP, and obtaining third-stage APPs related to the APPs to be expanded according to the APPs searched by the second-stage keywords in the application library platform; obtaining keywords covered by each third-stage APP, and obtaining a candidate keyword set according to the keywords covered by each third-stage APP;
the similarity calculation module is used for obtaining the similarity of each third-stage APP relative to the second-stage APP and obtaining 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 similarity and the proportion;
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 keyword expansion apparatus based on the associated application, because the content of information interaction, execution process, and the like between the modules is based on the same concept as that of the foregoing method embodiment of the present invention, the technical effect brought by the content is the same as that of the foregoing method embodiment of the present invention, and specific content may refer to the description in the method embodiment of the present invention, and is not described herein again.
In addition, in the above exemplary embodiment of the related application-based keyword expansion apparatus, 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 configuration requirements of corresponding hardware or convenience of implementation of software, that is, the internal structure of the related application-based keyword expansion apparatus 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. Which when executed 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, where the program is executed by a processor to implement any one of the above-mentioned keyword expansion methods based on associated applications.
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 further provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the keyword expansion method based on the associated application as in any one of the above embodiments when executing the program.
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 associated application is characterized by comprising the following steps:
acquiring first-stage keywords covered by the APP to be expanded, and obtaining second-stage APPs associated with the APPs to be expanded according to the APPs searched by the first-stage keywords in the application library platform;
acquiring second-stage keywords covered by each second-stage APP, and obtaining third-stage APPs related to the APPs to be expanded according to the APPs searched by each second-stage keyword in the application library platform; obtaining keywords covered by each third-stage APP, and obtaining a candidate keyword set according to the keywords covered by each third-stage APP;
determining the similarity of each third-stage APP relative to the second-stage APP, and 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 similarity and the proportion;
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 method for keyword expansion based on associated applications according to claim 1, wherein the obtaining of the first-level keywords covered by the APP to be expanded comprises:
acquiring all keywords covered by the APP to be expanded according to the historical search records of the application library platform; all keywords covered by the APP to be expanded are subjected to abnormal screening to delete the abnormal keywords in the keywords to be expanded, so that first-level keywords covered by the APP to be expanded are obtained;
and/or the presence of a gas in the gas,
the second-level keywords covered by the second-level APPs are obtained by the method comprising the following steps:
acquiring all keywords covered by each second-level APP according to the historical search record of the application library platform; performing exception screening on all keywords covered by each second-level APP to delete the abnormal keywords therein to obtain second-level keywords covered by the second-level APP;
and/or the presence of a gas in the gas,
obtaining keywords covered by each third-level APP comprises:
acquiring all keywords covered by each third-stage APP according to the historical search record of the application library platform; performing exception screening on all keywords covered by each third-stage APP to delete the abnormal keywords therein to obtain the keywords covered by the third-stage APP;
the abnormal keywords include: the search index is abnormal, the data of the keyword search result is abnormal, the APP ranks in the search result to be abnormal, and the keyword is at least one characteristic of the character number abnormality.
3. The keyword expansion method based on associated applications according to claim 2, wherein obtaining the second-level APP associated with the APP to be expanded according to the APP searched by each first-level keyword on the application library platform comprises:
obtaining the APP frequency sequencing information in the multiple search results corresponding to the first-stage keywords according to the multiple search results of the first-stage keywords in the historical search records within a set historical time period; acquiring a set number of APPs with a frequency sequence arranged in front as the APPs searched by each first-stage keyword;
obtaining an APP matrix according to all the first-level keywords and APP information searched by all the first-level keywords; 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;
and/or the presence of a gas in the gas,
the step of obtaining the third-level APP associated with the APP to be expanded according to the APPs searched by the second-level keywords on the application library platform includes:
obtaining the APP frequency sequencing information in the multiple search results corresponding to the second-level keywords according to the multiple search results of the second-level keywords in the historical search records within the set historical time period; acquiring a set number of APPs with a frequency sequence arranged in front as the APPs searched by each second-level keyword;
obtaining an APP matrix according to all the second-level keywords and the APPs searched by the second-level keywords; 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 second set frequency in the APP matrix as a third-stage APP associated with the second-stage APP.
4. The keyword expansion method based on associated applications as claimed in claim 3, wherein after obtaining the second-level APPs, before obtaining the second-level keywords covered by each second-level APP, further comprising:
acquiring an application list to which the APP to be expanded belongs in an application library platform, and deleting a second-level APP which belongs to a different application list from the APP to be expanded;
and/or the presence of a gas in the gas,
after obtaining the third-level APP, before obtaining the keywords covered by each third-level APP, the method further includes:
and acquiring an application list of the APP to be expanded in the application library platform, and deleting a third-stage APP belonging to different application lists with the APP to be expanded.
5. The keyword expansion method based on the associated application according to claim 1, wherein determining the similarity of each third-stage APP with respect to the second-stage APP comprises:
if one second-stage APP corresponding to the third-stage APP is available, acquiring the similarity between the third-stage APP and the corresponding second-stage APP as the similarity of the third-stage APP relative to the second-stage APP;
if the number of the second-stage APPs corresponding to the third-stage APP is more than two, the similarity between the third-stage APP and each corresponding second-stage APP is respectively obtained, so that a similarity mean value is calculated, and the similarity mean value is used as the similarity of the third-stage APP relative to the second-stage APPs.
6. The keyword expansion method based on associated application according to claim 5, wherein the obtaining of the similarity between the third-stage APP and the corresponding second-stage APP includes:
obtaining a feature vector of the second-stage APP according to the second-stage keywords covered by the second-stage APP, and obtaining a feature vector of each third-stage APP according to the keywords covered by each third-stage APP;
processing the feature vector of the second-stage APP and the feature vector of the third-stage APP through One-Hot coding to obtain a sparse feature vector of the second-stage APP and a sparse feature vector of the third-stage APP;
and calculating the similarity between each third-stage APP and the corresponding second-stage APP according to the sparse feature vector of the second-stage APP and the sparse feature vector of the third-stage APP.
7. The keyword expansion method based on associated application of claim 6, characterized in that the similarity between each third-level APP and the corresponding second-level APP is calculated by the following formula:
in the formula, APP (2) t Represents the t-th second-stage APP; s (3) i Represents the ith third-stage APP; v (APP) (2) t )·V(S (3) i ) Denotes APP (2) t Sparse feature vector of (1) and (S) (3) i Inner product of sparse feature vectors of (d); i V (APP) (2) t )|| 2 ||V(S (3) i )|| 2 Denotes APP (2) t Sparse feature vector of (1) and (S) (3) i Is the product of the 2-norm of the sparse feature vector.
8. The keyword expansion method based on associated applications as claimed in claim 7, wherein obtaining a candidate keyword set according to the keywords covered by each third-level APP comprises:
obtaining a keyword matrix according to third-stage APPs associated with the second-stage APPs and keywords covered by the third-stage APPs; merging and counting the keywords in the keyword matrix to obtain a candidate keyword set KW (3) =(kw (3) 1 ,kw (3) 2 ,…,kw (3) n ) And the corresponding keyword frequency vector is C (3) =(c 1 ,c 2 ,…,c n ) Each element of the keyword frequency vector respectively corresponds to the occurrence frequency of each keyword in the candidate keyword set;
the candidate keyword set KW (3) The proportion of the ith keyword in the list is as follows:
where i =1,2, …, n, n represents a candidate keyword set KW (3) The total number of keywords contained therein.
9. The method for keyword expansion based on associated application according to any one of claims 1 to 8, wherein the calculating a similarity score of each keyword in a candidate keyword set according to the similarity and the specific gravity includes:
obtaining a similarity score of the keywords in the candidate keyword set according to the product of the proportion of the keywords in the candidate keyword set and the similarity of the third-stage APP corresponding to the keywords relative to the second-stage APP;
and/or the presence of a gas in the gas,
screening the candidate keyword set according to the similarity score to obtain associated keywords of the APP to be expanded, wherein the screening comprises the following steps:
selecting a set number of keywords with the similarity scores ranked from high to low from a 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 associated application is characterized by comprising:
the application expansion module is used for acquiring first-stage keywords covered by the APP to be expanded, and obtaining second-stage APPs related to the APPs to be expanded according to the APPs searched by the first-stage keywords in the application library platform;
the candidate word expansion module is used for acquiring second-stage keywords covered by each second-stage APP, and obtaining third-stage APPs related to the APPs to be expanded according to the APPs searched by the second-stage keywords in the application library platform; obtaining keywords covered by each third-stage APP, and obtaining a candidate keyword set according to the keywords covered by each third-stage APP;
the similarity calculation module is used for determining the similarity of each third-stage APP relative to the second-stage APP and 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 similarity and the proportion;
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.
CN201711227933.4A 2017-11-29 2017-11-29 Keyword expanding method and device based on associated application Active CN108021640B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711227933.4A CN108021640B (en) 2017-11-29 2017-11-29 Keyword expanding method and device based on associated application

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711227933.4A CN108021640B (en) 2017-11-29 2017-11-29 Keyword expanding method and device based on associated application

Publications (2)

Publication Number Publication Date
CN108021640A true CN108021640A (en) 2018-05-11
CN108021640B CN108021640B (en) 2019-08-16

Family

ID=62077455

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711227933.4A Active CN108021640B (en) 2017-11-29 2017-11-29 Keyword expanding method and device based on associated application

Country Status (1)

Country Link
CN (1) CN108021640B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101246484A (en) * 2007-02-15 2008-08-20 刘二中 Electric text similarity processing method and system convenient for query
CN103488780A (en) * 2013-09-27 2014-01-01 中国联合网络通信集团有限公司 Application program searching method and device
CN103902720A (en) * 2014-04-10 2014-07-02 北京博雅立方科技有限公司 Method and device for acquiring expansion words of keywords
CN105426360A (en) * 2015-11-12 2016-03-23 中国建设银行股份有限公司 Keyword extracting method and device
CN106326300A (en) * 2015-07-02 2017-01-11 富士通株式会社 Information processing method and information processing device
CN106503224A (en) * 2016-11-04 2017-03-15 维沃移动通信有限公司 A kind of method and device for recommending application according to keyword

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101246484A (en) * 2007-02-15 2008-08-20 刘二中 Electric text similarity processing method and system convenient for query
CN103488780A (en) * 2013-09-27 2014-01-01 中国联合网络通信集团有限公司 Application program searching method and device
CN103902720A (en) * 2014-04-10 2014-07-02 北京博雅立方科技有限公司 Method and device for acquiring expansion words of keywords
CN106326300A (en) * 2015-07-02 2017-01-11 富士通株式会社 Information processing method and information processing device
CN105426360A (en) * 2015-11-12 2016-03-23 中国建设银行股份有限公司 Keyword extracting method and device
CN106503224A (en) * 2016-11-04 2017-03-15 维沃移动通信有限公司 A kind of method and device for recommending application according to keyword

Also Published As

Publication number Publication date
CN108021640B (en) 2019-08-16

Similar Documents

Publication Publication Date Title
CN108320171B (en) Hot-sold commodity prediction method, system and device
CN110866181B (en) Resource recommendation method, device and storage medium
CN111798273A (en) Training method of purchase probability prediction model of product and purchase probability prediction method
CN106445963B (en) Advertisement index keyword automatic generation method and device of APP platform
US11514498B2 (en) System and method for intelligent guided shopping
CN108182200B (en) Keyword expansion method and device based on semantic similarity
CN109582155B (en) Recommendation method and device for inputting association words, storage medium and electronic equipment
CN111159563A (en) Method, device and equipment for determining user interest point information and storage medium
CN111091883B (en) Medical text processing method, device, storage medium and equipment
CN111160699A (en) Expert recommendation method and system
CN110968802A (en) User characteristic analysis method, analysis device and readable storage medium
CN109522275B (en) Label mining method based on user production content, electronic device and storage medium
CN108170664B (en) Key word expansion method and device based on key words
CN110955774B (en) Word frequency distribution-based character classification method, device, equipment and medium
CN108170665B (en) Keyword expansion method and device based on comprehensive similarity
JP4891638B2 (en) How to classify target data into categories
CN112182264A (en) Method, device and equipment for determining landmark information and readable storage medium
CN106708880B (en) Topic associated word acquisition method and device
CN115827990B (en) Searching method and device
CN109462635B (en) Information pushing method, computer readable storage medium and server
CN108052554A (en) The method and apparatus that various dimensions expand keyword
CN108021640A (en) Keyword expanding method and device based on associated application
JP2016162127A (en) Image search apparatus, method and program
CN112069388B (en) Entity recommendation method, system, computer device and computer readable storage medium
CN108182201B (en) Application expansion method and device based on key keywords

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