CN105159884B - The method for building up and device of industry dictionary and industry recognition methods and device - Google Patents
The method for building up and device of industry dictionary and industry recognition methods and device Download PDFInfo
- Publication number
- CN105159884B CN105159884B CN201510613993.4A CN201510613993A CN105159884B CN 105159884 B CN105159884 B CN 105159884B CN 201510613993 A CN201510613993 A CN 201510613993A CN 105159884 B CN105159884 B CN 105159884B
- Authority
- CN
- China
- Prior art keywords
- search
- industry
- search term
- dictionary
- sub
- 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
Links
Abstract
The method for building up and device of a kind of industry dictionary of offer of the embodiment of the present invention and industry recognition methods and device.The method for building up of the sector dictionary includes：Obtain user's search behavior daily record；Each search term and the corresponding search result being clicked are extracted from user's search behavior daily record；The affiliated industry of search term is determined according to the search result being clicked, using described search word as accurate index word, establishes and preserves the accurate index word and the entry pair of corresponding industry, form accurate dictionary.Analysis based on user's search behavior daily record can establish accurate dictionary automatically, since the search result being clicked reflects the search need of user, determine the affiliated industry of search term based on this, the accuracy rate of obtained entry pair is high；Since corresponding all search results being clicked to accurate index word are analyzed, the omission of certain accurate index words and the one-to-many correspondence of industry is avoided, improves the accuracy rate of the entry pair of accurate dictionary.
Description
Technical field
The present embodiments relate to information discriminating technology field more particularly to a kind of method for building up and device of industry dictionary
And industry recognition methods and device.
Background technology
The existing industry identification for search behavior is based primarily upon manually generated vocabulary, only when search term hits vocabulary
When, the identification of industry could be carried out.
It has a drawback in that：Manually generated vocabulary is low for the coverage rate of search term；For the multiple rows of certain correspondences
The special search term of industry, using manually generated vocabulary, each search term only corresponds to an industry, causes industry identification accurate
True rate is relatively low.
Invention content
The embodiment of the present invention provides a kind of method for building up and device of industry dictionary, to realize the automatic Building of industry dictionary
It is vertical.
The embodiment of the present invention also provides a kind of industry recognition methods and device, to improve the coverage rate to query string, and carries
The accuracy rate of the industry identification of high query string.
In a first aspect, an embodiment of the present invention provides a kind of method for building up of industry dictionary, including：
Obtain user's search behavior daily record；
Each search term and the corresponding search result being clicked are extracted from user's search behavior daily record；
The affiliated industry of search term is determined according to the search result being clicked, using described search word as accurate index
Word is established and preserves the accurate index word and the entry pair of corresponding industry, forms accurate dictionary.
Second aspect establishes device an embodiment of the present invention provides a kind of industry dictionary, including：
Log acquisition module, for obtaining user's search behavior daily record；
Extraction module, for extracting from user's search behavior daily record each search term and corresponding being clicked
Search result；
Accurate dictionary forms module, and the search result for being clicked according to determines the affiliated industry of search term, by institute
Search term is stated as accurate index word, establishes and preserves the accurate index word and the entry pair of corresponding industry, formed accurate
Dictionary.
The third aspect, an embodiment of the present invention provides a kind of industry recognition methods, are provided based on any embodiment of the present invention
The dictionary established of method for building up of industry dictionary realize, including：
Obtain query string input by user；
The query string is accurately matched in the accurate dictionary pre-established, it will be corresponding to the accurate index word of successful match
Industry as the corresponding industry of the query string, and return to the corresponding industry of the query string.
Fourth aspect, an embodiment of the present invention provides a kind of industry identification devices, are provided based on any embodiment of the present invention
The dictionary that device established of establishing of industry dictionary realize, including：
Query string acquisition module, for obtaining query string input by user；
Industry identification module, for accurately matching the query string in the accurate dictionary pre-established, by successful match
Accurate index word corresponding to industry as the corresponding industry of the query string, and return to the corresponding industry of the query string.
The method for building up and device of industry dictionary provided in an embodiment of the present invention, point based on user's search behavior daily record
Analysis, can automatically set up accurate dictionary, with the update of user's search behavior daily record, can constantly update in accurate dictionary
Entry pair improves coverage rate of the accurate dictionary to search term；That usually reflected due to the search result being clicked is user
Search need, therefore the affiliated industry of search term is determined according to the search result being clicked, the accuracy rate of obtained entry pair is high；Together
When, for each accurate index word in accurate dictionary, due to corresponding all search knots being clicked to accurate index word
Fruit is analyzed, and avoids the omission of certain accurate index words and the one-to-many correspondence of industry, improves accurate word
The accuracy rate of the entry pair of allusion quotation.
Industry recognition methods provided in an embodiment of the present invention and device, since accurate dictionary is based on user's search behavior day
What the analysis of will was automatically set up, with the update of user's search behavior daily record, the entry in accurate dictionary can be constantly updated
It is right, improve coverage rate of the accurate dictionary to query string；Using accurate dictionary, the industry identification to query string can be realized, especially
It is the query string higher to search rate industry identification, either one-one relationship accurately identify or one
Many-many relationship accurately identifies, and improves the accuracy rate of the industry identification to query string.
Description of the drawings
In order to illustrate more clearly of the present invention, one will be done to attached drawing needed in the present invention below and be simply situated between
It continues, it should be apparent that, the accompanying drawings in the following description is some embodiments of the present invention, and those of ordinary skill in the art are come
It says, without having to pay creative labor, can also be obtained according to these attached drawings other attached drawings.
Fig. 1 a are the flow diagram of the method for building up of a kind of industry dictionary that the embodiment of the present invention one provides；
Fig. 1 b are the search knot that a kind of basis is clicked in the method for building up of industry dictionary that the embodiment of the present invention one provides
Fruit shape is into the method flow schematic diagram of accurate dictionary；
Fig. 2 is the structure diagram for establishing device of a kind of industry dictionary that the embodiment of the present invention four provides；
Fig. 3 is the flow diagram of a kind of industry recognition methods that the embodiment of the present invention five provides；
Fig. 4 is the structure diagram of a kind of industry identification device that the embodiment of the present invention six provides.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached drawing to the embodiment of the present invention
In technical solution be described in further detail, it is clear that described embodiment is part of the embodiment of the present invention rather than complete
The embodiment in portion.It is understood that specific embodiment described herein is only used for explaining the present invention rather than to the present invention's
It limits, based on the embodiments of the present invention, those of ordinary skill in the art are obtained without creative efforts
Every other embodiment, shall fall within the protection scope of the present invention.It also should be noted that for ease of description, attached drawing
In illustrate only part related to the present invention rather than full content.
It should be mentioned that some exemplary embodiments are described as before exemplary embodiment is discussed in greater detail
The processing described as flow chart or method.Although operations (or step) are described as the processing of sequence by flow chart,
Many of which operation can be implemented concurrently, concomitantly or simultaneously.In addition, the sequence of operations can be pacified again
Row.The processing can be terminated when its operations are completed, it is also possible to have the additional step being not included in attached drawing.Institute
It states processing and can correspond to method, function, regulation, subroutine, subprogram etc..
It should further be mentioned that in some replaces realization modes, the function/action being previously mentioned can be according to different from attached
The sequence indicated in figure occurs.For example, depending on involved function/action, each width figure shown in succession actually may be used
Substantially simultaneously to perform or can perform in a reverse order sometimes.
Embodiment one
A is please referred to Fig.1, the flow diagram of the method for building up of a kind of industry dictionary provided for the embodiment of the present invention one.
The method of the embodiment of the present invention can be by being configured establishing device and perform with hardware and/or the industry dictionary of software realization, should
Realization device is typically configured in the server for being capable of providing information search service.
This method includes：Operation 110~operation 130.
110th, user's search behavior daily record is obtained.
User various search engines scan for be a kind of common acquisition information by way of.User's searches each time
Rope operates, and search engine can all form corresponding user's search behavior daily record, not only including search term input by user, further include
Corresponding search result and user are to information such as the clicks of certain specific search results.
The source of user's search behavior daily record is not limited in this operation, either user's search of mobile terminal
User behaviors log or user's search behavior daily record at PC ends.Further, for mobile terminal and/or PC ends, both
Can be the corresponding use of a certain vertical channel classification search (for example, " webpage " this vertical classification is searched in Baidu search engine)
Family search behavior daily record can also be multiple vertical channel classification search (for example, in Baidu search engine " webpage " and " map "
The search of two vertical classifications) corresponding user's search behavior daily record, user's search behavior daily record of full platform can also be expanded to.
In this operation, the user search behavior daily record of (such as 3 month) is typically obtained in setting time.
120th, each search term and the corresponding search result being clicked are extracted from user's search behavior daily record.
130th, the search result being clicked according to determines the affiliated industry of search term, using described search word as accurate rope
Draw word, establish and preserve the accurate index word and the entry pair of corresponding industry, form accurate dictionary.
In the multiple search process for same search term, since the search need of different user is different, lead to what is had
Search result is clicked, and some search results are not clicked, and that usually reflected due to the search result being clicked is user
Search need, therefore the affiliated industry of search term is determined according to the search result being clicked, based on this, obtained entry pair it is accurate
Rate is high.
For some search terms, it can only determine that these search terms only correspond to an industry according to the search result being clicked；
And for other search term, these search terms while corresponding multiple industries can be determined according to the search result being clicked,
Therefore, for the entry in accurate dictionary to there are two types of form, one kind is one-to-one entry pair, a kind of is one-to-many entry pair.
For example, it is assumed that search term is " KFC ", the search result being clicked has：Include searching for KFC shops information
Rope as a result, include the search result and the search result for including KFC group purchase information that KFC makes a reservation on the net, according to
These search results being clicked, it may be determined that industry belonging to search term " KFC " is " cuisines ", therefore can be obtained accurate
The entry pair of index terms " KFC " and corresponding industry " cuisines " is one-to-one entry pair, is stored in accurate dictionary.
For another example, it is assumed that search term is " exhibition center ", and the search result being clicked has：Include the Exhibition Information in museum
The search knot of search result, the search result for including the Exhibition Information of the Art Museum and the Exhibition Information for including science and technology center
Fruit, according to the search result for the Exhibition Information for including museum being clicked, it may be determined that row belonging to search term " exhibition center "
Industry is " museum ", according to the search result for the Exhibition Information for including the Art Museum being clicked, it may be determined that search term " exhibition
Industry belonging to shop " is " Art Museum ", according to the search result for the Exhibition Information for including science and technology center being clicked, it may be determined that search
Industry belonging to rope word " exhibition center " is " science and technology center ", therefore can obtain accurate index word " exhibition center " and corresponding industry and " win
The entry pair in object shop ", " Art Museum " and " science and technology center " is one-to-many entry pair, is stored in accurate dictionary.
The technical solution of the present embodiment based on the analysis of user's search behavior daily record, can automatically set up accurate dictionary,
With the update of user's search behavior daily record, the entry pair in accurate dictionary can be constantly updated, improves accurate dictionary to searching
The coverage rate of rope word；Due to the search result being clicked usually reflect be user search need, according to what is be clicked
Search result determines the affiliated industry of search term, and the accuracy rate of obtained entry pair is high；Meanwhile for each essence in accurate dictionary
True index terms since corresponding all search results being clicked to accurate index word are analyzed, avoids certain
The omission of accurate index word and the one-to-many correspondence of industry improves the accuracy rate of the entry pair of accurate dictionary.
As the search result that basis is clicked, a kind of preferred embodiment of accurate dictionary is formed, please refers to Fig.1 b,
It may particularly include：Operation 131~operation 132.
131st, the search rate of each search term and the click probability of the corresponding search result being clicked are counted.
It is preceding to have addressed, typically obtain user's search behavior daily record in setting time.In setting time, user often into
Row one search operation, then total searching times add 1, it is corresponding total can to count to obtain user's search behavior daily record in setting time
Searching times, and the searching times of each search term are counted, for each search term, according to the searching times of the search term and total search
Number, you can statistics obtains the search rate of the search term.
In the multiple search process for same search term, since the search need of different user is different, lead to what is had
The number of clicks of search result is higher, and the number of clicks of some search results is relatively low, and search result also is not clicked, and is passed through
The click probability of the corresponding search result being clicked of each search term is counted, is conducive to know the search need of user.
132nd, it is more than each search term of the first threshold value for described search frequency, is clicked the search term is corresponding
When the click probability of search result is more than the second threshold value, using the search term as accurate index word, and determine corresponding described
Industry of the probability belonging to more than the search result of the second threshold value is clicked, as the corresponding industry of the accurate index word；It establishes
And the accurate index word and the entry pair of corresponding industry are preserved, form accurate dictionary.
For example, it is assumed that the search rate of search term " KFC " is more than the first threshold value, include KFC shops information
The click probability of search result be more than the second threshold value, include the click probability of the search result that KFC is made a reservation on the net
More than the second threshold value, include the click probability of search result of KFC group purchase information also greater than the second threshold value, pass through
It determines, the industry belonging to these three search results is " cuisines ", therefore using search term " KFC " as accurate index word, will
" cuisines " it is " beautiful with corresponding industry to obtain accurate index word " KFC " as the corresponding industry of accurate index word " KFC "
The entry pair of food ", is stored in accurate dictionary.
Similarly, the one-to-one entry pair in accurate dictionary can be obtained, e.g., accurate index word " China Merchants Bank " with
The entry pair of corresponding industry " bank ".
For another example, it is assumed that the search rate of search term " ABC hotels " is more than the first threshold value, and include the hotel moves in letter
The click probability of the search result of breath is more than the second threshold value, includes the point of the search result of the dish information of hotel offer
Probability is hit also greater than the second threshold value, by determining, the industry belonging to the search result for the check-in information for including the hotel is
" hotel ", the industry belonging to the search result for the dish information for including hotel offer is " cuisines ", therefore by search term
" ABC hotels " as accurate index word, by " hotel " and " cuisines " as the corresponding industry of accurate index word " ABC hotels ",
Accurate index word " ABC hotels " and corresponding industry " hotel " and the entry pair of " cuisines " are obtained, is stored in accurate dictionary.
Similarly, the one-to-many entry pair in accurate dictionary can be obtained, e.g., accurate index word " exhibition center " with it is corresponding
Industry " museum ", " Art Museum " and " science and technology center " entry pair.
The technical solution of the present embodiment based on the analysis of user's search behavior daily record, can automatically set up accurate dictionary,
With the update of user's search behavior daily record, the entry pair in accurate dictionary can be constantly updated, improves accurate dictionary to searching
The coverage rate of rope word；For each entry pair in accurate dictionary, since the search rate of accurate index word is more than the first thresholding
Value so that the higher search term of search rate is encompassed by accurate dictionary, is conducive to the row of the higher search term of search rate
Industry identifies, and has carried out statistical analysis, and will click on to the click probability of corresponding each search result of accurate index word
Probability is more than the industry belonging to each search result of the second threshold value, is determined as the corresponding industry of accurate index word, avoids certain
The omission of a little accurate index words and the one-to-many correspondence of industry improves the accuracy rate of the entry pair of accurate dictionary.
Embodiment two
The present embodiment provides a kind of method for building up of industry dictionary, the present embodiment is being united on the basis of above-described embodiment
After the search rate for counting each search term, further include：
It is less than or equal to each search term of the first threshold value for described search frequency, using the accurate dictionary, this is searched
Rope word is split, and obtains the corresponding sub- search term of the search term and the corresponding industry of the sub- search term；By the search term
Corresponding sub- search term is established and preserves the fuzzy index word and the sub- search term is corresponding as fuzzy index word
The entry pair of industry forms fuzzy dictionary.
For example, it is assumed that the search rate of search term " neighbouring China Merchants Bank " is less than or equal to the first threshold value, using accurate
Dictionary splits the search term, obtains the corresponding sub- search term " China Merchants Bank " of the search term and sub- search term " is recruited
The corresponding industry of quotient bank " " bank ", therefore, by the corresponding sub- search term " China Merchants Bank " of search term " neighbouring China Merchants Bank "
As word is searched for generally, establish and search for word " China Merchants Bank " and the entry pair of industry " bank " generally, be stored in fuzzy dictionary.
Similarly, the one-to-one entry pair in fuzzy dictionary can be obtained, e.g., fuzzy index word " KFC " with it is right
The entry pair of industry " cuisines " answered.
It should be noted that certain entries pair in accurate dictionary with the fuzzy entry in fuzzy dictionary to can partly weigh
It is folded, such as " KFC " → " cuisines " this entries pair, it can both appear in inside accurate dictionary, while mould can also be appeared in
It pastes inside dictionary, but for accurate dictionary, search term and accurate index word " Ken De in user's search behavior daily record
Base " accurately matches, and specific acquisition modes above-described embodiment of entry pair has been addressed, repeated no more；For fuzzy dictionary, use
Search term in the search behavior daily record of family is with searching for word " KFC " fuzzy matching generally, for example, user's search term is " neighbouring
KFC ", acquisition modes are different from the acquisition modes of the entry pair in accurate dictionary, are that search term is carried out using accurate dictionary
Split what is obtained.
For another example, it is assumed that the search rate of search term " China Merchants Bank near KFC " is less than or equal to the first threshold value, profit
With accurate dictionary, which is split, obtains the corresponding sub- search term " KFC " of the search term and " China Merchants Bank ",
And the sub- corresponding industry " cuisines " of search term " KFC " and the corresponding industry " bank " of sub- search term " China Merchants Bank ", because
This, the corresponding sub- search term " KFC " of search term " China Merchants Bank near KFC " and " China Merchants Bank " are searched as fuzzy
Rope word, establishes entry pair：" KFC ", " China Merchants Bank " → " cuisines ", " bank ", are stored in fuzzy dictionary.
The technical solution of the present embodiment, on the one hand, the analysis based on user's search behavior daily record can automatically set up essence
True dictionary with the update of user's search behavior daily record, can constantly update the entry pair in fuzzy dictionary, improve accurate word
Allusion quotation is to the coverage rate of search term；On the other hand, for each entry pair in accurate dictionary, due to the search frequency of accurate index word
Rate be more than the first threshold value so that the higher search term of search rate is encompassed by accurate dictionary, be conducive to search rate compared with
The industry identification of high search term；In another aspect, to the click probability of corresponding each search result of accurate index word into
It has gone statistical analysis, and will click on industry of the probability belonging to more than each search result of the second threshold value, be determined as accurate index
The corresponding industry of word avoids the omission of certain accurate index words and the one-to-many correspondence of industry, improves accurate dictionary
Entry pair accuracy rate；Another aspect,, can be with using accurate dictionary based on the further analysis of user's search behavior daily record
Automatically set up fuzzy dictionary, for the relatively low search term of search rate (such as long-tail search term), using accurate dictionary without
When method accurately matches, fuzzy matching is carried out using fuzzy dictionary, can realize the row of the relatively low search term of this part searches frequency
Industry identifies, the coverage rate to search term is further improved by accurate dictionary and fuzzy dictionary.
Embodiment three
The present embodiment provides a kind of method for building up of industry dictionary, the present embodiment is being formed on the basis of embodiment two
After fuzzy dictionary, further include：
The corresponding sub- search term of the search term quantity be at least two when, determine strategy using priority, determine to
The priority of few two sub- search terms, establishes and preserves described at least two sub- search terms and the sub- search term pair of highest priority
The entry pair for the industry answered forms priority dictionary.
It is still illustrated by taking the search term " China Merchants Bank near KFC " in above-described embodiment two as an example, this is searched
Rope word, it is preceding to have addressed, using accurate dictionary, entry pair can be obtained：" KFC ", " China Merchants Bank " → " cuisines ", " bank ",
It is stored in fuzzy dictionary.The corresponding sub- search term of search term " China Merchants Bank near KFC " is two, is " Ken De respectively
Base " and " China Merchants Bank ", using priority determine strategy (there are many strategy, behind can be described in detail), determine the two
The priority of sub- search term is：The priority of sub- search term " China Merchants Bank " is higher than the priority of sub- search term " KFC ", and
The corresponding industry of sub- search term " China Merchants Bank " of the highest priority is " bank ", therefore, establishes entry pair：" KFC ",
" China Merchants Bank " → " bank " forms priority dictionary.
The technical solution of the present embodiment based on the analysis of user's search behavior daily record, can establish accurate dictionary, base automatically
In the further analysis of user's search behavior daily record, and accurate dictionary is utilized, fuzzy dictionary can be established automatically, wherein, it utilizes
Accurate dictionary can realize the industry identification of the search term higher to search rate, either the accurate knowledge of one-one relationship
Not or many-one relationship accurately identifies, improve the search term higher to search rate industry identification it is accurate
Rate；For the relatively low search term of search rate (such as long-tail search term), when can not accurately be matched using accurate dictionary, utilize
Fuzzy dictionary carries out fuzzy matching, can realize the industry identification of the relatively low search term of this part searches frequency, pass through accurate word
Allusion quotation and fuzzy dictionary further improve the coverage rate to search term, wherein, for the relatively low search term of certain search rates, profit
With entry pair one-to-one in fuzzy dictionary, the corresponding industry of these search terms can recognize that, the accuracy rate of industry identification is higher
(for example, for the relatively low search term of search rate " neighbouring China Merchants Bank ", utilize the one-to-one entry in fuzzy dictionary
To " China Merchants Bank " → " bank ", the industry that can recognize the search term is " bank ")；For other search rate compared with
Low search term needs the entry using multi-to-multi in fuzzy dictionary to carrying out industry identification, correspondingly, can identify to obtain multiple
Corresponding industry at this point, using priority dictionary, is searched due to considering the corresponding multiple sons of the relatively low search term of the search rate
The priority of rope word, and the priority between multiple sub- search terms is conducive to the search need for reflecting user, therefore utilizes fuzzy
Use of the entry of multi-to-multi to cooperation priority dictionary in dictionary can improve the accuracy rate of industry identification (for example, for searching
The relatively low search term of rope frequency " China Merchants Bank near KFC ", using the multi-to-multi entry in fuzzy dictionary to " Ken De
Base ", " China Merchants Bank " → " cuisines ", " bank ", identification obtain the corresponding industry of the search term as " cuisines " and " bank ", utilize
Entry in priority dictionary can identify to obtain the corresponding row of the search term to " KFC ", " China Merchants Bank " → " bank "
Industry is " bank ").
In said program, strategy is determined using priority, determines this operation of the priority of at least two sub- search terms
It can be there are many realization method.
Position that can be according to described at least two sub- search terms in corresponding search term determines the son search of position rearward
The priority of the priority of the word search term forward higher than position；Such as search term " China Merchants Bank near KFC " is corresponding
Sub- search term for two, be " KFC " and " China Merchants Bank " respectively, the position of neutron search term " China Merchants Bank " rearward,
The position of sub- search term " KFC " is forward, it is thus determined that the priority of the two sub- search terms is：Sub- search term " silver of promoting trade and investment
The priority of row " is higher than the priority of sub- search term " KFC ".
Also can be long according to the words of described at least two sub- search terms, determine that word is grown the priority of longer sub- search term and is higher than
Word grows the priority of shorter sub- search term；Such as the corresponding sub- search term of search term " China Merchants Bank near KFC " is two
It is a, it is " KFC " and " China Merchants Bank " respectively, the word of neutron search term " China Merchants Bank " grows longer, sub- search term " Ken De
The word length of base " is shorter, it is thus determined that the priority of the two sub- search terms is：The priority of sub- search term " China Merchants Bank " is higher than
The priority of sub- search term " KFC ".
Can also according to the click probability of the search result being clicked of the corresponding search term of described at least two sub- search terms,
Determine the priority for clicking the higher corresponding sub- search term of the search result being clicked of the probability quilt relatively low higher than probability is clicked
The priority of the corresponding sub- search term of search result of click.For example, search term " China Merchants Bank near KFC " is corresponding
The search result being clicked is through counting as two classes comprising having the click probability of the search result of China Merchants Bank far above packet
The click probability of search result containing KFC, accordingly, it is determined that the priority of the two sub- search terms is：Sub- search term " is recruited
The priority of quotient bank " is higher than the priority of sub- search term " KFC ".
It should be noted that by determining that the priority between multiple sub- search terms is conducive to reflect that the search of user needs
It asks, for example, when user searches for " China Merchants Bank near KFC ", really desired is search knot about China Merchants Bank
Therefore fruit rather than the search result about KFC, for the relatively low search term of certain search rates, need to utilize fuzzy word
The entry of multi-to-multi is to carrying out industry identification in allusion quotation, correspondingly, can identify to obtain multiple corresponding industries, at this point, using preferential
Grade dictionary, due to considering the priority of the corresponding multiple sub- search terms of the relatively low search term of the search rate, and multiple sons are searched
Priority between rope word is conducive to the search need for reflecting user, therefore utilizes the entry of multi-to-multi in fuzzy dictionary to cooperation
The use of priority dictionary can improve the accuracy rate of industry identification.
Above-mentioned several realization methods can be individually performed, and can also combine use, and the present embodiment is not limited this.
It should also be noted that, for being individually performed, between multiple sub- search terms determined by three kinds of realization methods
The accuracy rate of priority is different, is followed successively by from high to low：It is excellent according to the click determine the probability of the corresponding search result of sub- search term
The mode of first grade>According to the mode of sub- search term location determination priority>The mode for determining priority is grown according to sub- search term word.
Example IV
Referring to Fig. 2, a kind of structure diagram for establishing device of industry dictionary provided for the embodiment of the present invention four.It should
Device includes：Log acquisition module 210, extraction module 220 and accurate dictionary form module 230.
Wherein, log acquisition module 210 is used to obtain user's search behavior daily record；Extraction module 220 is used for from the use
Each search term and the corresponding search result being clicked are extracted in the search behavior daily record of family；Accurate dictionary forms module 230 and uses
The affiliated industry of search term is determined in the search result being clicked according to, using described search word as accurate index word, is established
And the accurate index word and the entry pair of corresponding industry are preserved, form accurate dictionary.
In said program, the accurate dictionary, which forms module, to be specifically used for：The search rate of each search term is counted, with
And the click probability of the corresponding search result being clicked；It is more than each search of the first threshold value for described search frequency
Word, the corresponding search result being clicked of the search term click probability be more than the second threshold value when, using the search term as
Accurate index word, and determine the corresponding industry for clicking probability belonging to more than the search result of the second threshold value, as institute
State the corresponding industry of accurate index word；It establishes and preserves the accurate index word and the entry pair of corresponding industry, formed accurate
Dictionary.
Further, described device may also include：
Fuzzy dictionary forms module, small for described search frequency for after the search rate for counting each search term
In each search term equal to the first threshold value, using the accurate dictionary, which is split, obtains the search term pair
The sub- search term and the corresponding industry of the sub- search term answered；Using the corresponding sub- search term of the search term as fuzzy index
Word is established and preserves the fuzzy index word and the entry pair of the corresponding industry of the sub- search term, forms fuzzy dictionary.
Further, described device may also include：Priority Determination module and priority dictionary form module.
Wherein, Priority Determination module is used for after fuzzy dictionary is formed, in the corresponding sub- search term of the search term
When quantity is at least two, determines strategy using priority, determine the priority of at least two sub- search terms；Priority dictionary shape
It is used to establish and preserve the word of described at least two sub- search terms industry corresponding with the sub- search term of highest priority into module
Item pair forms priority dictionary.
Specifically, the Priority Determination module may include following at least one submodules：First priority determines submodule
Block, the second priority determination sub-module and third priority determination sub-module.
Wherein, the first priority determination sub-module is used for according to described at least two sub- search terms in corresponding search term
Position, determine the priority of the priority of the sub- search term of the position rearward sub- search term forward higher than position；Second is preferential
Word of the grade determination sub-module for according to described at least two sub- search terms is long, determines that word grows the priority of longer sub- search term
The priority of shorter sub- search term is grown higher than word；Third priority determination sub-module is used to be searched for according to described at least two sons
The click probability of the search result being clicked of the corresponding search term of word determines to click the higher search result being clicked of probability
The priority of corresponding sub- search term, which is higher than, clicks the preferential of the relatively low corresponding sub- search term of the search result being clicked of probability
Grade.
The device of establishing of industry dictionary provided in an embodiment of the present invention can perform the row that any embodiment of the present invention is provided
The method for building up of industry dictionary has the corresponding function module of execution method and advantageous effect.
Embodiment five
Referring to Fig. 3, a kind of flow diagram of industry recognition methods provided for the embodiment of the present invention five.It is of the invention real
The method for applying example can be performed by the industry identification device realized with hardware and/or software is configured, and the realization device is typical
It is arranged in the server for being capable of providing the identification service of information industry.The method of the present embodiment is based on any embodiment of the present invention
The dictionary that the method for building up of the industry dictionary of offer is established is realized.
This method includes：Operation 310~operation 320.
310th, query string input by user is obtained.
320th, the query string is accurately matched in the accurate dictionary pre-established, by the accurate index word institute of successful match
Corresponding industry returns to the corresponding industry of the query string as the corresponding industry of the query string.
The technical solution of the present embodiment, since accurate dictionary is that the analysis based on user's search behavior daily record is automatically set up
, with the update of user's search behavior daily record, the entry pair in accurate dictionary can be constantly updated, improves accurate dictionary pair
The coverage rate of query string；Using accurate dictionary, the industry identification to query string can be realized, it is especially higher to search rate
The industry identification of query string, either accurately identifying for one-one relationship or accurately identifying for many-one relationship, carry
The high accuracy rate identified to the industry of query string.
In said program, after accurately matching the query string in the accurate dictionary pre-established, the method is also
It may include：
If it fails to match, the query string described in fuzzy matching in the fuzzy dictionary pre-established, by successful match
The corresponding industry of fuzzy index word returns to the corresponding industry of the query string as the corresponding industry of the query string.
The method for the relatively low query string of search rate (such as long-tail search term), is utilizing accurate dictionary smart
Really during matching, fuzzy matching is carried out using fuzzy dictionary, can realize that the industry of the relatively low query string of this part searches frequency is known
, coverage rate to query string is not further improved by accurate dictionary and fuzzy dictionary, wherein, for certain search rates compared with
Low query string using entry pair one-to-one in fuzzy dictionary, can recognize that the corresponding industry of these query strings, industry identification
Accuracy rate it is higher.
Further, in the fuzzy dictionary pre-established after query string described in fuzzy matching, the method can also wrap
It includes：
When the quantity for the fuzzy index word for detecting successful match is at least two, the priority word pre-established is utilized
Allusion quotation determines the corresponding industry of at least two fuzzy indexes word, as the corresponding industry of the query string, and is looked into described in return
Ask the corresponding industry of string.
The method for the relatively low query string of search rate (such as long-tail search term), is utilizing accurate dictionary smart
Really during matching, fuzzy matching is carried out using fuzzy dictionary, can realize that the industry of the relatively low query string of this part searches frequency is known
, coverage rate to query string is not further improved by accurate dictionary and fuzzy dictionary, wherein, for certain search rates compared with
Low query string needs the entry using multi-to-multi in fuzzy dictionary to carrying out industry identification, correspondingly, can identify to obtain multiple
Corresponding industry, at this point, using priority dictionary, due to considering the corresponding multiple sub- search of the relatively low query string of search rate
The priority of word, and the priority between multiple sub- search terms is conducive to the search need for reflecting user, therefore utilizes fuzzy word
Use of the entry of multi-to-multi to cooperation priority dictionary in allusion quotation can improve the accuracy rate of industry identification.
To sum up, the method for present embodiments providing the corresponding industry of three kinds of identification query strings, one kind are to utilize accurate word
Allusion quotation carries out industry identification, and another kind is in the case of using accurate dictionary recognition failures, and industry knowledge is carried out using fuzzy dictionary
Not, another is using in the case of accurate dictionary recognition failures, using fuzzy dictionary and coordinates the use of priority dictionary,
Realize industry identification.
In above-mentioned various industries recognition methods, after the corresponding industry of the query string is returned, the method may be used also
Including：
According to the corresponding industry of the query string, a pair search result corresponding with the query string carries out truncation, obtains
Result is recalled to the query string is corresponding；
Result is recalled described in return.
This programme can be applied to the backstage of a variety of vertical channel classification search, for filtered search as a result, obtaining recalling knot
Fruit.Such as the backstage applied to vertical channel classification search " Baidu map ", for recalling and blocking.Application mode is to pass through
After industry identification, the search of general demand will carry out matching in itself according to query string and recall, can also be according to the industry of identification into every trade
Industry is recalled, and recalling result can be blocked according to the result that industry identifies, so as to influence to recall as a result, recognition result also may be used simultaneously
For being ranked up to recalling result.
In above-mentioned various industries recognition methods, after the corresponding industry of the query string is returned, the method may be used also
Including：
Information recommendation is carried out according to the corresponding industry of the query string；
Or including：
According to the corresponding industry of the query string, recommendation information is determined, and choose component exhibiting；
The recommendation information is handled according to selected component exhibiting, returns to the handling result.
Both schemes can be applied to the foreground of a variety of vertical channel classification search.To be applied to vertical channel classification search
For the foreground of " Baidu map ", the information recommendation according to industry recognition result and displaying are briefly introduced.For example, it is assumed that
User input inquiry string " China Merchants Bank " in the search box of " Baidu map ", the industry recognized is " bank ", according to the row
Industry recognition result determines corresponding recommendation information, and chooses component exhibiting corresponding with industry " bank " so that recommendation information is pressed
Be shown according to the component exhibiting, e.g., show recommendation information in the form of a list, and show the business hours of bank, be ATM also
It is the information such as agency.For another example, it is assumed that user's input inquiry string " KFC " in the search box of " Baidu map " recognizes
Industry is " cuisines ", according to the sector recognition result, determines corresponding recommendation information, and choose exhibition corresponding with industry " cuisines "
Show component so that recommendation information is shown according to the component exhibiting, e.g., show each KFC shops group purchase information, whether
Support makes a reservation, whether collects the information such as the expense sent outside on the net.For another example, it is assumed that user's input inquiry in the search box of " Baidu map "
It goes here and there " Four Seasons Hotel ", the industry recognized is " hotel ", according to the sector recognition result, determines corresponding recommendation information, and select
Take component exhibiting corresponding with industry " hotel " so that recommendation information is shown according to the component exhibiting, e.g., shows each four seasons
The address in hotel is moved in and the information such as subscribes.
Embodiment six
Referring to Fig. 4, a kind of structure diagram of industry identification device provided for the embodiment of the present invention six.The present embodiment
The dictionary that device established of establishing of device industry dictionary based on the embodiment of the present invention realize.The device includes：It looks into
Ask string acquisition module 410 and industry identification module 420.
Wherein, query string acquisition module 410 is used to obtain query string input by user；Industry identification module 420 is used for
Accurately match the query string in the accurate dictionary pre-established, using the industry corresponding to the accurate index word of successful match as
The corresponding industry of the query string, and return to the corresponding industry of the query string.
In said program, the industry identification module 420 can be additionally used in the accurate dictionary pre-established accurate
After the query string, if it fails to match, the query string described in fuzzy matching in the fuzzy dictionary pre-established, general
With the corresponding industry of successful fuzzy index word as the corresponding industry of the query string, and return to the corresponding row of the query string
Industry.
Further, the industry identification module 420 can be additionally used in the fuzzy matching institute in the fuzzy dictionary pre-established
It is excellent using what is pre-established when the quantity for the fuzzy index word for detecting successful match is at least two after stating query string
First grade dictionary determines the corresponding industry of at least two fuzzy indexes word, as the corresponding industry of the query string, and returns
The corresponding industry of the query string.
In said program, described device may also include：It recalls result acquisition module and recalls result return module.
Wherein, it recalls result acquisition module to be used for after the corresponding industry of the query string is returned, according to the inquiry
Go here and there corresponding industry, a pair search result corresponding with the query string carries out truncation, obtains that the query string is corresponding to call together
Return result；It recalls result return module and described recalls result for returning.
In said program, described device may also include：
Information recommendation module, it is corresponding according to the query string for after the corresponding industry of the query string is returned
Industry carries out information recommendation；
Or including：Recommend element determining module and displaying processing module.
Wherein, element determining module is recommended to be used for after the corresponding industry of the query string is returned, according to the inquiry
It goes here and there corresponding industry, determines recommendation information, and choose component exhibiting；Show that processing module is used for according to selected component exhibiting
The recommendation information is handled, returns to the handling result.
Industry identification device provided in an embodiment of the present invention can perform the industry identification that any embodiment of the present invention is provided
Method has the corresponding function module of execution method and advantageous effect.
Finally it should be noted that：The above various embodiments is merely to illustrate technical scheme of the present invention rather than it is limited
System；Preferred embodiment in embodiment, is not intended to limit it, to those skilled in the art, the present invention can be with
There are various modifications and changes.All any modification, equivalent replacement, improvement and so within spirit and principles of the present invention,
It should be included within protection scope of the present invention.
Claims (14)
1. a kind of method for building up of industry dictionary, which is characterized in that including：
Obtain user's search behavior daily record；
Each search term and the corresponding search result being clicked are extracted from user's search behavior daily record；
Count the search rate of each search term and the click probability of the corresponding search result being clicked；
It is more than each search term of the first threshold value for search rate, in the point of the corresponding search result being clicked of the search term
When hitting probability more than the second threshold value, using the search term as accurate index word, and determine that the corresponding click probability is more than
Industry belonging to the search result of second threshold value, as the corresponding industry of the accurate index word；It establishes and preserves the essence
The entry pair of true index terms and corresponding industry, forms accurate dictionary；
It is less than or equal to each search term of the first threshold value for described search frequency, using the accurate dictionary, to the search term
It is split, obtains the corresponding sub- search term of the search term and the corresponding industry of the sub- search term；The search term is corresponded to
Sub- search term as fuzzy index word, establish and preserve the fuzzy index word and the corresponding industry of the sub- search term
Entry pair, form fuzzy dictionary.
2. according to the method described in claim 1, it is characterized in that, after fuzzy dictionary is formed, the method further includes：
When the quantity of the corresponding sub- search term of the search term is at least two, strategy is determined using priority, determines at least two
The priority of a sub- search term, is established and to preserve described at least two sub- search terms corresponding with the sub- search term of highest priority
The entry pair of industry forms priority dictionary.
3. according to the method described in claim 2, it is characterized in that, determine strategy using priority, determine that at least two sons are searched
The priority of rope word, including at least one of following：
According to position of the described at least two sub- search terms in corresponding search term, the excellent of the sub- search term of position rearward is determined
The priority of the first grade sub- search term forward higher than position；
It is long according to the word of described at least two sub- search terms, determine that the priority of the longer sub- search term of word length is shorter higher than word length
Sub- search term priority；
According to the click probability of the search result being clicked of the corresponding search term of described at least two sub- search terms, determine to click
The priority of the corresponding sub- search term of the higher search result being clicked of probability higher than click probability it is relatively low be clicked search
The priority of the corresponding sub- search term of hitch fruit.
4. a kind of industry dictionary establishes device, which is characterized in that including：
Log acquisition module, for obtaining user's search behavior daily record；
Extraction module, for extracting each search term and the corresponding search being clicked from user's search behavior daily record
As a result；
Accurate dictionary forms module, for counting the search rate of each search term and the corresponding search knot being clicked
The click probability of fruit；
It is more than each search term of the first threshold value for described search frequency, in the corresponding search result being clicked of the search term
Click probability when being more than the second threshold value, using the search term as accurate index word, and the determining corresponding click probability
More than the industry belonging to the search result of the second threshold value, as the corresponding industry of the accurate index word；It establishes and preserves institute
Accurate index word and the entry pair of corresponding industry are stated, forms accurate dictionary；
Fuzzy dictionary forms module, for after the search rate for counting each search term, being less than for described search frequency etc.
In each search term of the first threshold value, using the accurate dictionary, which is split, it is corresponding to obtain the search term
Sub- search term and the corresponding industry of the sub- search term；Using the corresponding sub- search term of the search term as fuzzy index word, build
It stands and preserves the fuzzy index word and the entry pair of the corresponding industry of the sub- search term, form fuzzy dictionary.
5. device according to claim 4, which is characterized in that described device further includes：
Priority Determination module, for being extremely in the quantity of the corresponding sub- search term of the search term after fuzzy dictionary is formed
When two few, determine strategy using priority, determine the priority of at least two sub- search terms；
Priority dictionary forms module, and the son for establishing and preserving described at least two sub- search terms and highest priority is searched for
The entry pair of the corresponding industry of word forms priority dictionary.
6. device according to claim 5, which is characterized in that the Priority Determination module includes following at least one sons
Module：
First priority determination sub-module, for the position according to described at least two sub- search terms in corresponding search term,
Determine the priority of the priority of the sub- search term of the position rearward sub- search term forward higher than position；
Second priority determination sub-module, it is long for the word according to described at least two sub- search terms, determine that word grows longer son
The priority of search term grows the priority of shorter sub- search term higher than word；
Third priority determination sub-module, for according to the corresponding search term of described at least two sub- search terms be clicked search
The click probability of hitch fruit, the priority for determining to click the higher corresponding sub- search term of the search result being clicked of probability are higher than
Click the priority of the relatively low corresponding sub- search term of the search result being clicked of probability.
7. a kind of industry recognition methods, the word that the method for building up based on any industry dictionaries of claim 1-3 is established
Allusion quotation is realized, which is characterized in that including：
Obtain query string input by user；
The query string is accurately matched in the accurate dictionary pre-established, by the row corresponding to the accurate index word of successful match
Industry returns to the corresponding industry of the query string as the corresponding industry of the query string；
If it fails to match, the query string described in fuzzy matching in the fuzzy dictionary pre-established, by the fuzzy of successful match
The corresponding industry of index terms returns to the corresponding industry of the query string as the corresponding industry of the query string.
8. it the method according to the description of claim 7 is characterized in that is looked into described in fuzzy matching in the fuzzy dictionary pre-established
After asking string, the method further includes：
When the quantity for the fuzzy index word for detecting successful match is at least two, using the priority dictionary pre-established,
It determines the corresponding industry of at least two fuzzy indexes word, as the corresponding industry of the query string, and returns to the inquiry
Go here and there corresponding industry.
It is 9. described according to the method described in claim 8, it is characterized in that, after the corresponding industry of the query string is returned
Method further includes：
According to the corresponding industry of the query string, a pair search result corresponding with the query string carries out truncation, obtains institute
State that query string is corresponding to recall result；
Result is recalled described in return.
It is 10. described according to the method described in claim 8, it is characterized in that, after the corresponding industry of the query string is returned
Method further includes：
Information recommendation is carried out according to the corresponding industry of the query string；Or
According to the corresponding industry of the query string, recommendation information is determined, and choose component exhibiting；According to selected component exhibiting
The recommendation information is handled, returns to the handling result.
11. a kind of industry identification device, the word established device and established based on any industry dictionaries of claim 4-6
Allusion quotation is realized, which is characterized in that including：
Query string acquisition module, for obtaining query string input by user；
Industry identification module, for accurately matching the query string in the accurate dictionary pre-established, by the essence of successful match
Industry corresponding to true index terms returns to the corresponding industry of the query string as the corresponding industry of the query string；Also use
After the query string is accurately matched in the accurate dictionary pre-established, if it fails to match, in the mould pre-established
Query string described in fuzzy matching in dictionary is pasted, the corresponding industry of fuzzy index word of successful match is corresponded to as the query string
Industry, and return to the corresponding industry of the query string.
12. according to the devices described in claim 11, which is characterized in that the industry identification module is additionally operable to pre-establishing
Fuzzy dictionary in after query string described in fuzzy matching, be at least two in the quantity for the fuzzy index word for detecting successful match
When a, using the priority dictionary pre-established, determine the corresponding industry of at least two fuzzy indexes word, looked into as described
The corresponding industry of string is ask, and returns to the corresponding industry of the query string.
13. device according to claim 12, which is characterized in that described device further includes：
Result acquisition module is recalled, it is corresponding according to the query string for after the corresponding industry of the query string is returned
Industry, pair corresponding with a query string search result carry out truncation, obtain that the query string is corresponding to recall result；
Result return module is recalled, described result is recalled for returning.
14. device according to claim 12, which is characterized in that described device further includes：
Information recommendation module, for after the corresponding industry of the query string is returned, according to the corresponding industry of the query string
Carry out information recommendation；
Or：
Recommend element determining module, it is corresponding according to the query string for after the corresponding industry of the query string is returned
Industry determines recommendation information, and chooses component exhibiting；
It shows processing module, for being handled according to selected component exhibiting the recommendation information, returns to the processing
As a result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510613993.4A CN105159884B (en) | 2015-09-23 | 2015-09-23 | The method for building up and device of industry dictionary and industry recognition methods and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510613993.4A CN105159884B (en) | 2015-09-23 | 2015-09-23 | The method for building up and device of industry dictionary and industry recognition methods and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105159884A CN105159884A (en) | 2015-12-16 |
| CN105159884B true CN105159884B (en) | 2018-06-29 |
Family
ID=54800743
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510613993.4A Active CN105159884B (en) | 2015-09-23 | 2015-09-23 | The method for building up and device of industry dictionary and industry recognition methods and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105159884B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107423362B (en) * | 2017-06-20 | 2021-06-08 | 创新先进技术有限公司 | Industry determining method, object obtaining method and device, client and server |
| CN107612707B (en) * | 2017-08-04 | 2021-04-09 | 深圳市其乐游戏科技有限公司 | Preprocessing method and system for classified storage of homologous sample data in industry field |
| CN107832468B (en) * | 2017-11-29 | 2019-05-10 | 百度在线网络技术（北京）有限公司 | Demand recognition methods and device |
| CN108536800B (en) * | 2018-04-03 | 2022-04-19 | 有米科技股份有限公司 | Text classification method, system, computer device and storage medium |
| CN109885753A (en) * | 2019-01-16 | 2019-06-14 | 苏宁易购集团股份有限公司 | A kind of method and device for expanding commercial articles searching and recalling |
| CN110688558A (en) * | 2019-09-10 | 2020-01-14 | 中国平安财产保险股份有限公司 | Method and device for searching web page, electronic equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101140587A (en) * | 2007-10-15 | 2008-03-12 | 深圳市迅雷网络技术有限公司 | Searching method and apparatus |
| CN102169495A (en) * | 2011-04-11 | 2011-08-31 | 趣拿开曼群岛有限公司 | Industry dictionary generating method and device |
| CN103136339A (en) * | 2013-02-01 | 2013-06-05 | 百度在线网络技术（北京）有限公司 | Searching method, client-side and network server-side based on service information |
| CN103226601A (en) * | 2013-04-25 | 2013-07-31 | 百度在线网络技术（北京）有限公司 | Method and device for image search |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014047727A1 (en) * | 2012-09-28 | 2014-04-03 | Alkis Papadopoullos | A method and system for monitoring social media and analyzing text to automate classification of user posts using a facet based relevance assessment model |
-
2015
- 2015-09-23 CN CN201510613993.4A patent/CN105159884B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101140587A (en) * | 2007-10-15 | 2008-03-12 | 深圳市迅雷网络技术有限公司 | Searching method and apparatus |
| CN102169495A (en) * | 2011-04-11 | 2011-08-31 | 趣拿开曼群岛有限公司 | Industry dictionary generating method and device |
| CN103136339A (en) * | 2013-02-01 | 2013-06-05 | 百度在线网络技术（北京）有限公司 | Searching method, client-side and network server-side based on service information |
| CN103226601A (en) * | 2013-04-25 | 2013-07-31 | 百度在线网络技术（北京）有限公司 | Method and device for image search |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105159884A (en) | 2015-12-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105159884B (en) | The method for building up and device of industry dictionary and industry recognition methods and device | |
| WO2017121251A1 (en) | Information push method and device | |
| US8103658B2 (en) | Index backbone join | |
| US8583685B2 (en) | Determination of category information using multiple stages | |
| CN107590267B (en) | Information-pushing method and device, terminal and readable storage medium storing program for executing based on picture | |
| CN104216992B (en) | A kind of information processing method and device | |
| US20140172821A1 (en) | Generating filters for refining search results | |
| US11003700B2 (en) | Methods and systems for query segmentation in a search | |
| CN104750839B (en) | A kind of data recommendation method, terminal and server | |
| CN103368992A (en) | Message push method and device | |
| CN110727862B (en) | Method and device for generating query strategy of commodity search | |
| CN103186666B (en) | Method, device and equipment for searching based on favorites | |
| CN109657044A (en) | Data retrieval method, data reordering method, device, terminal and storage medium | |
| CN112463859B (en) | User data processing method and server based on big data and business analysis | |
| CN110532351A (en) | Recommend word methods of exhibiting, device, equipment and computer readable storage medium | |
| CN103955480B (en) | A kind of method and apparatus for determining the target object information corresponding to user | |
| CN105745642B (en) | Inquire the treating method and apparatus of data | |
| CN111159563A (en) | Method, device and equipment for determining user interest point information and storage medium | |
| CN109543113B (en) | Method and device for determining click recommendation words, storage medium and electronic equipment | |
| CN108959453A (en) | Information extracting method, device and readable storage medium storing program for executing based on text cluster | |
| CN102053960B (en) | Method and system for constructing quick and accurate Internet of things and Internet search engine according to group requirement characteristics | |
| CN104408144B (en) | The detection method and device of web search keyword | |
| CN106934679A (en) | information matching method and device | |
| CN109426974A (en) | Competing product analysis method and device | |
| CN107818152B (en) | Plant retrieval method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |