CN103810198A - Method and device for searching commodity information - Google Patents
Method and device for searching commodity information Download PDFInfo
- Publication number
- CN103810198A CN103810198A CN201210448776.0A CN201210448776A CN103810198A CN 103810198 A CN103810198 A CN 103810198A CN 201210448776 A CN201210448776 A CN 201210448776A CN 103810198 A CN103810198 A CN 103810198A
- Authority
- CN
- China
- Prior art keywords
- search
- word
- participle
- words
- point
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a method and a device for searching commodity information, which are used for solving the problem of more pressure of a server caused by inaccurate search results provided by the server in the prior art. According to the method, all commodity types corresponding to a search word input by a user are determined, so that the preferential commodity types of the user are selected, and a search word is determined again based on the search word input by the user and headwords of all commodities under all the selected commodity types and is used for searching. By adopting the method, the search word input by the user is reconstructed based on the search word input by the user and the headwords of the commodities under the preferential commodity types of the user, and the reconstructed search word not only contains the search intention of the user this time but also contains the preference of the user, so that the accuracy of the obtained search results can be effectively improved by searching based on the reconstructed search word, and thus the pressure of the server is reduced.
Description
Technical field
The application relates to communication technical field, relates in particular to a kind of searching method and device of merchandise news.
Background technology
At present, most of shopping websites all provide function of search to user, and by function of search, user can input to search word the server of website, and the server of website is searched for relevant merchandise news and returns to user according to the search word of user's input.
In the prior art, in order to improve the matching degree of the actual intention of Search Results and user, to avoid, user is follow-up still to be needed by continuous change search word or screening conditions is set to carry out repeat search, cause the excessive problem of server stress, server is in the time carrying out the search of merchandise news, conventionally need to determine user's preference, then search in conjunction with the search word of user's input and user's preference.Concrete, server, in the time carrying out the search of merchandise news, first extracts the each merchandise news relevant to user's preference, then adopts the search word that user inputs to search in the each merchandise news extracting, and finally Search Results is returned to user.
But, user's preference is to be added up and obtain according to the operation behavior information in this user's past by server, can only characterize haply this user's preference, and only with regard to Client-initiated is once searched for, the search intention of current search just often has stronger provisional and singularity, and it wants the merchandise news of search probably not follow the preference in the past in this user.For example, certain user's that server is determined preference is the shirt of brand A, but the search word of this user's input is brand B, be not its brand A of preference in the past, adopt method of the prior art, server will first extract the merchandise news relevant to the shirt of brand A, then adopts search word " brand B " to search in the merchandise news extracting.Obviously, the merchandise news that server extracts is all the merchandise news relevant to the shirt of brand A, the Search Results that the more difficult acquisition of the merchandise news user that search is relevant to brand B in such merchandise news needs.
Therefore, the method that the search word of available technology adopting user's input is searched in each merchandise news relevant to preference this user that extract, may not provide Search Results accurately for user, even may search for less than Search Results, cause user still can or screening conditions be set by continuous change search word and carry out repeat search, cause server stress excessive.
Summary of the invention
The embodiment of the present application provides a kind of searching method and device of merchandise news, inaccurate in order to solve the Search Results that in prior art, server provides, and causes the problem that server stress is larger.
The searching method of a kind of merchandise news that the embodiment of the present application provides, comprising:
Determine each commodity classification corresponding to search word that user inputs;
Preference degree value according to predetermined described user to each commodity classification is selected commodity classification in each commodity classification of determining;
Search word to described user's input carries out word segmentation processing, obtain each search participle, heading to the commodity class of selecting all commodity now carries out word segmentation processing, obtain each title participle, each search participle based on obtaining and each title participle, build point word combination that includes at least one search participle and at least one title participle;
According to point word combination of setting up in advance and the corresponding relation of standard words, determine respectively standard words corresponding to each point of word combination building, wherein, the search word that standard words adopts while comprising each search in the past, for a standard words, the each participle having in point word combination of corresponding relation with this standard words is all participles of this standard words;
Respectively a point set of words gathered to form in the search participle and the title participle that in each point of word combination of the same standard words of correspondence, comprise;
Redefine search word according to the search participle comprising in point set of words and title participle;
The search word that employing redefines is searched for, and Search Results is returned to described user.
The searcher of a kind of merchandise news that the embodiment of the present application provides, comprising:
Search classification determination module, for determining each commodity classification corresponding to search word of user's input;
Preference classification determination module for the preference weights to each commodity classification according to predetermined described user, is selected commodity classification in each commodity classification of determining;
Divide word combination determination module, for the search word of described user's input is carried out to word segmentation processing, obtain each search participle, heading to the commodity class of selecting all commodity now carries out word segmentation processing, obtain each title participle, each search word based on obtaining and each title participle, build point word combination that includes at least one search participle and at least one title participle;
Standard words determination module, point word combination of setting up in advance for basis and the corresponding relation of standard words, determine respectively standard words corresponding to each point of word combination building, wherein, the search word that standard words adopts while comprising each search in the past, for a standard words, the each participle having in point word combination of corresponding relation with this standard words is all participles of this standard words;
Set determination module, gathers to form a point set of words for the search participle and the title participle that respectively each point of word combination of the same standard words of correspondence are comprised;
Search word reconstructed module, redefines search word for the search participle and the title participle that comprise according to point set of words;
Search module, for adopting the search word redefining to search for, and returns to described user by Search Results.
The embodiment of the present application provides a kind of searching method and device of merchandise news, the method is determined each commodity classification corresponding to search word that user inputs, therefrom select the commodity classification of user institute preference, respectively the search word of user's input and the heading of the commodity class of selecting all commodity are now carried out to participle, obtain each search participle and title participle, structure includes point word combination of at least one search participle and title participle, determine the each point of standard words that word combination is corresponding, determine in the each point of word combination by corresponding identical standard word and comprise point set of words that participle forms, and redefine search word according to the participle comprising in point set of words, the search word that employing redefines is searched for.The heading of the search word that said method is inputted based on user and the commodity class of user preference commodity now, search word to user's input has carried out reconstruct, the search word reconstructing had both comprised the search intention of this search of user, comprise again user's preference, therefore the search word based on reconstruct is searched for, can effectively improve the accuracy of the Search Results obtaining, thereby reduce the pressure of server.
Accompanying drawing explanation
The search procedure of the merchandise news that Fig. 1 provides for the embodiment of the present application;
The searcher structural representation of the merchandise news that Fig. 2 provides for the embodiment of the present application.
Embodiment
While determining user's preference due to server in prior art, be to determine according to the operation behavior information in user's past, therefore definite preference is this user preference haply, and the search intention of user while at every turn initiating to search for often has stronger provisional and singularity, it wants the merchandise news of search probably not follow in preference in the past, therefore, the method that adopts search word that user inputs to search in prior art in each merchandise news relevant to this user's preference, often can not provide Search Results accurately for user, even may search for less than any Search Results, thereby cause user still can or screening conditions be set by continuous change search word and carry out repeat search, cause server stress excessive.
In the embodiment of the present application in order to improve the accuracy of search, to reduce the pressure of server, in the time that user initiates to search for, server is first determined each commodity classification corresponding to search word of user's input, therefrom select again the commodity classification of user preference, the heading of the search word based on user's input and the commodity class of user preference commodity now, is reconstructed the search word of user's input, finally adopts the search word of reconstruct to search for.Because the search word reconstructing had both comprised the search intention of this search of user, comprise again user's preference, therefore the search word based on reconstruct is searched for, and can effectively improve the accuracy of the Search Results obtaining, thereby reduces the pressure of server.
Below in conjunction with Figure of description, the embodiment of the present application is described in detail.
The search procedure of the merchandise news that Fig. 1 provides for the embodiment of the present application, specifically comprises the following steps:
S101: determine each commodity classification corresponding to search word that user inputs.
In the embodiment of the present application, server can be set up the corresponding relation of search word and commodity classification in advance.
Concrete, the method that server is set up the corresponding relation of search word and commodity classification comprises: the search that server carried out for the past at every turn, determine the search word adopting while carrying out this search, determine and adopt this search word to search for, and Search Results is returned to after the user of this search word of input, input the commodity classification under the merchandise news that the user of this search word clicks in Search Results, set up this commodity classification of determining and the corresponding relation of this search word.
For example, suppose that the search word that once search adopts carrying out is in the past " brand A ", server is in the time carrying out this search, Search Results is returned to after the user of this search word of input (" brand A "), input the shirt that has comprised brand A in the merchandise news that the user of this search word clicks in Search Results, and because the commodity classification under the shirt of this brand A is shirt classification, therefore, server is set up the corresponding relation of this search word " brand A " and shirt classification.Here the object that, user clicks is also the merchandise news that user clicks.Certainly, user also may click many merchandise newss, if the merchandise news of clicking also comprises the traveling bag of this brand A, and commodity classification under the traveling bag of this brand A is case and bag classification, also can be using case and bag classification as search word commodity classification corresponding to " brand A ".
That is to say, in the corresponding relation of the search word of setting up and commodity classification, a search word can corresponding multiple commodity classifications.For example, suppose that server set up the corresponding relation of search word " brand B " and classification " shirt ", classification " fitted pants ", classification " western-style clothes ", classification " leather shoes " in advance, if the search word of user's input is " brand B ", server determines that the commodity classification corresponding to search word of this user's input is classification " shirt ", classification " fitted pants ", classification " western-style clothes ", classification " leather shoes ", and above-mentioned four commodity classifications of determining are the commodity classification relevant to the search word " brand B " of this user's input.
Preferably, in order further to improve the accuracy of subsequent searches, server is set up in the process of corresponding relation of search word and commodity classification at employing said method, at the search word adopting for certain search of carrying out in the past, determined input this search word user click merchandise news under commodity classification after, the degree of relevancy value of this commodity classification and this search word can also be increased to default step value, and in the time that the degree of relevancy value of determining current this commodity classification and this search word is not less than setting dependent thresholds, set up again the corresponding relation of this commodity classification and this search word.
For example, the search word adopting in the once search of carrying out is in the past " brand A ", server determines that the merchandise news that the user of this search word of input (" brand A ") clicks in Search Results is the merchandise news of brand A shirt, also be, the affiliated commodity classification of merchandise news of inputting user's click of this search word is shirt classification, the degree of relevancy value of this shirt classification and this search word " brand A " can be increased to the step value of setting, for example, add 1.If user all belongs to shirt classification in many merchandise newss of the click of Search Results, by the step value of setting, the degree of relevancy value of shirt classification and this search word " brand A " is increased successively.Whether the degree of relevancy value that judges current this shirt classification and this search word " brand A " is not less than setting dependent thresholds, if, set up the corresponding relation of this search word " brand A " and shirt classification, otherwise, first temporarily do not set up the corresponding relation of this search word " brand A " and shirt classification, set dependent thresholds until the degree of relevancy value of this shirt classification and this search word " brand A " is not less than.
Understandable, for the less commodity classification of click volume in the Search Results under search word, can be not as the corresponding commodity classification of this search word.Now, the method of setting up the corresponding relation of search word and commodity classification can also comprise: the click that counting user carries out under this search word occurs in each commodity class number of times now, filtering number of clicks is lower than the commodity classification of setting threshold, and after carrying out filtering step, remaining commodity classification is as commodity classification corresponding to this search word.
Server adopts said method to set up in advance after the corresponding relation of each search word and each commodity classification, in the time receiving the search word of user's input, just can be according to the search word of setting up in advance and the corresponding relation of commodity classification, determine each commodity classification corresponding to search word of this user's input, each commodity classification of determining is the commodity classification relevant to the search word of this user's input, and also this initiates the relevant commodity classification of search intention of search to this user.
Understandably, in the embodiment of the present application, also commodity classification that can be under each object in the Search Results under the search word of user input is as commodity classification corresponding to this search word, and need not consider the historical click data of user to the Search Results under this search word.
S102: the preference degree value according to predetermined this user to each commodity classification, in each commodity classification of determining, select commodity classification.
In the embodiment of the present application, server has pre-determined the preference degree value of this user to each commodity classification, this user can characterize the preference degree of this user to this commodity classification to the preference degree value of a commodity classification, preference degree value is larger, illustrate that this user is higher to the preference degree of this commodity classification, otherwise, illustrate that this user is lower to the preference degree of this commodity classification.Concrete, server determines that the method for the preference degree value of this user to each commodity classification comprises: server is for each commodity classification, according to the assigned operation that this user's past carries out this commodity class commodity now, determine the preference degree value of this user to this commodity classification.Wherein, assigned operation comprises: one or more in browse operation, order operation, search operation, certainly, also can be according to actual needs using the operation of other kinds as assigned operation.
Further, server is at the assigned operation of commodity now of certain commodity class being carried out according to this user's past, and while determining this user to the preference degree value of this commodity classification, number of times or the frequency of the assigned operation that specifically can carry out according to user be determined.For example, the every search of this user is this commodity class merchandise news now once, just this user is added to 1 to the preference degree value of this commodity classification, often browse once this commodity class merchandise news now, just this user is added to 2 to the preference degree value of this commodity classification, every order is this commodity class commodity now once, just this user are added to 5 to the preference degree value of this commodity classification.Certainly, also can adopt other algorithms to determine the preference degree value of this user to this commodity classification, the embodiment of the present application is not specifically limited the method for determining preference degree value.
Because server has pre-determined the preference degree value of this user to each commodity classification, therefore server can, accordingly in each commodity classification of determining by step S101, be selected the commodity classification of user preference.Concrete, server can, according to this user to the preference degree value of each commodity classification order from big to small, in each commodity classification of determining by step S101, be selected the commodity classification of setting quantity, as the commodity classification of this user preference successively.Wherein, setting quantity can set as required.Also can set preference degree threshold value, preference degree value is exceeded to the commodity classification of this preference degree threshold value as the commodity classification corresponding to search word of user's input.
For example, the search word of user's input is " brand B ", commodity classification corresponding to search word of determining this user's input is shirt classification, fitted pants classification, western-style clothes classification, leather shoes classification, suppose that predetermined this user of server is followed successively by from big to small to the preference degree value of above-mentioned four commodity classifications: shirt classification, western-style clothes classification, leather shoes classification, fitted pants classification, and setting quantity is 2, two commodity classifications that server is selected according to said sequence are shirt classification and leather shoes classification, these two commodity classifications are the classification of this user preference of further selecting each commodity classification that server is relevant from the search word " brand B " to this user input determined by step S101.
S103: the search word to this user's input carries out word segmentation processing, obtain each search participle, heading to the commodity class of selecting all commodity now carries out word segmentation processing, obtain each title participle, each search participle based on obtaining and each title participle, build point word combination that includes at least one search participle and at least one title participle.
In the embodiment of the present application, after server is further selected the commodity classification of user preference from each commodity classification relevant to the search word of this user's input, the heading of the search word of respectively this user being inputted and the commodity class of selecting by step S102 all commodity now carries out word segmentation processing, search word is carried out to word segmentation processing and obtain searching for participle, the heading of the commodity class of selecting commodity is now carried out to word segmentation processing and obtain title participle.
Then, the server each search participle based on obtaining and each title participle structure point word combination again.Wherein, in a point of word combination of formation, need to comprise at least one search participle and at least one title participle, and the quantity of the total participle comprising in a point of word combination should be not less than 2.
For example, suppose the search word Q of this user's input to carry out word segmentation processing, the search participle obtaining is q1 and q2, heading to the commodity class of selecting by step S102 some commodity now carries out word segmentation processing, and the title participle obtaining is t1, t2, t3, if form a point word combination in the mode that comprises altogether two participles in point word combination, point word combination building comprises: [q1, t1], [q1, t2], [q1, t3], [q2, t1], [q2, t2], [q2, t3], 6 totally points of word combinations.Similarly, for the heading of the commodity class of selecting by step S102 each commodity now, carry out word segmentation processing, and build a point word combination with search participle.
Can be found out by upper example, because point word combination building has comprised a search participle and a title participle, and search participle is the search word of user's input to be carried out to word segmentation processing obtain, can characterize user's search intention, title participle is the heading of the commodity class of user preference commodity now to be carried out to word segmentation processing obtain, can characterize user's preference, therefore, in point word combination building, both comprised this search intention of initiating search of user, comprise again user's preference, thereby, follow-up the participle combination of determining is processed and searched for, can improve the accuracy of search.
S104: according to point word combination of setting up in advance and the corresponding relation of standard words, determine respectively standard words corresponding to each point of word combination building.
In the embodiment of the present application, server has been set up the corresponding relation of point word combination and standard words in advance, wherein, the search word that standard words adopts while including but not limited to over each search, and, for a standard words, set up in advance to have with this standard words the each participle comprising in point word combination of corresponding relation are all participles of this standard words.Concrete, the method that server is set up the corresponding relation of point word combination and standard words comprises: the search word adopting when the past is searched for is at every turn defined as standard words, for each standard words, this standard words is carried out to word segmentation processing, obtain each participle of this standard words, point word combination that the participle of definite this standard words by any amount forms, sets up the corresponding relation that divides word combination and this standard words being made up of the participle of this standard words.Wherein, server can be set up a co-occurrence dictionary, and with each point of word combination falling that the mode of row chain sets up and the corresponding relation of standard words, the corresponding relation of foundation is kept in this co-occurrence dictionary.
For example, carry out in the past the search word that three whens search adopt and be respectively: A, B, C, using these three search words all as standard words, respectively these three standard words are carried out after word segmentation processing, the participle of A is that (a, b, c), the participle of B is that (a, b, d, e), the participle of C is (a, f).
Suppose that still the mode to comprise altogether two participles in point word combination forms a point word combination, for standard words A, because the participle of A is that (a, b, c), the word combination that divides being therefore made up of the participle of A has 3, is respectively [a, b], [a, c], [b, c].Similarly, for standard words B, point word combination being made up of the participle of B has 6, is respectively [a, b], [a, d], [a, e], [b, d], [b, e], [d, e]; For standard words C, point word combination being made up of the participle of C only has 1, is [a, f].
Based on the above-mentioned corresponding point word combination obtaining that three standard words (A, B, C) are processed, point word combination of foundation and the corresponding relation of standard words can be as shown in table 1.
Table 1
In above-mentioned table 1,3 points of word combinations that are made up of the participle of standard words A are all corresponding with standard words A, and 6 points of word combinations that are made up of the participle of standard words B are all corresponding with standard words B, and 1 point of word combination being made up of the participle of standard words C is all corresponding with standard words C.
And, as can be seen from Table 1, although a point of word combination is to be made up of the participle of same standard words, and in different standard words, may comprise two or more identical participles, as above the standard words A in example and standard words B all comprise participle a and participle b, therefore, same point of word combination may be corresponding different standard words, as above in example, divide word combination [a, b] corresponding to standard words A, also corresponding to standard words B.
If the corresponding relation with each point of word combination falling that the mode of row chain sets up with standard words, the corresponding relation of foundation can be as shown in table 2.
| Divide word combination | Corresponding standard words |
| [a，b] | A、B |
| [a，c] | A |
| [b，c] | A |
| [a，d] | B |
| [a，e] | B |
| [b，d] | B |
| [b，e] | B |
| [d，e] | B |
| [a，f] | C |
Table 2
By setting up corresponding relation as shown in table 2 to fall the mode of row chain, can facilitate the follow-up point corresponding standard words of word combination forming by step S103 of determining, to improve follow-up search efficiency.
Because server has been preserved after the corresponding relation of each point of word combination as shown in table 1 or table 2 and standard words in advance, therefore for point word combination (the search participle of the search word of being inputted by user building by above-mentioned steps S103, and the title participle of the heading of the commodity class of user preference commodity now forms), can determine the each point of standard words that word combination is corresponding.
S105: respectively a point set of words gathered to form in the search participle and the title participle that comprise in each point of word combination of the same standard words of correspondence.
In the each point of word combination building by step S103, for the each point of word combination corresponding to same standard words, think that these all participles that comprise corresponding to each point of word combination of same standard words (comprising search participle and title participle) have cooccurrence relation, therefore, for the each point of word combination corresponding to same standard words, determine point set of words that the search participle that comprises in these each point of word combinations corresponding to same standard words and title participle form, the search participle and the title participle that in point set of words forming, comprise are exactly the participle with cooccurrence relation.
S106: redefine search word according to the search participle comprising in point set of words and title participle.
In the embodiment of the present application, can, according to the participle with cooccurrence relation comprising in point set of words forming by step S105, redefine search word, also the search word of this user's input is reconstructed.Due to more than one of point set of words possibility forming by above-mentioned steps S105, therefore can from the each point of set of words of determining, select optimal set according to preset rules, again according to the search word of the search participle comprising in optimal set and this user's input of title participle reconstruct, concrete, the all participles that comprise in optimal set (comprising all search participles and title participle) can be spliced, reconstruct search word.
Further, whether the accuracy of considering Search Results depends primarily on relevant to the search word of this this input of user, and the quantity of the participle comprising in search word is more, the Search Results that subsequent searches arrives is also just more accurate, therefore, in the time of the each participle Resource selection optimal set from forming by step S105, can from the each point of set of words of determining, filter out point set of words that comprises all search participles, and in the each point of set of words filtering out, select point set of words that the quantity of the participle that comprises is maximum as optimal set.
In the optimal set that employing said method is selected, include all search participles of the search word of this this input of user, therefore when the search word that the participle that following adopted comprises based on this optimal set reconstructs is searched for, the Search Results searching must be the Search Results relevant to the search word of this this input of user, and because the quantity of the participle comprising in the optimal set of selecting is maximum, the accuracy of the Search Results that therefore subsequent searches arrives is also just very high, can effectively improve the accuracy of search, reduce the pressure of server.
In addition, consider that a search word or a standard words are to be made up of several participles, but form the competency difference to some extent of each participle of this search word or standard words, therefore, also can in server, set in advance the weights of expressing the meaning of each participle.The weights of expressing the meaning of a participle can characterize the competency of this participle, and the weights of expressing the meaning are higher, illustrate that the competency of this participle is higher, otherwise, illustrate that the competency of this participle is lower.Thereby, server is in the time of the each participle Resource selection optimal set from forming by step S105, also can from the each point of set of words of determining, filter out the set that comprises all search participles, and for the each point of set of words filtering out, according to the weights of expressing the meaning of setting for each participle in advance, determine all participles that comprise in this point of set of words filtering out the weights of expressing the meaning and value, in the each point of set of words filtering out, select to determine with point set of words of value maximum as optimal set.The each participle comprising in the optimal set that can guarantee like this to determine has higher competency, follow-up all participle reconstruct search words based on comprising in this optimal set, and search word based on reconstruct is while searching for, also can improve the accuracy of search.
S107: adopt the search word redefining to search for, and Search Results is returned to this user.
After the search word of this this input of user being reconstructed by above-mentioned steps S106, all search participles of the search word of this this input of user in search word due to reconstruct, are comprised, therefore the search intention that gives expression to this this search of user that the search word of reconstruct can be complete, and the title participle of the commodity class that has also comprised this user preference in the search word of reconstruct commodity now, also therefore the search word of reconstruct also can give expression to this user's preference, also be, the search word of reconstruct has incorporated the search intention of this this search of user and this user's preference, thereby adopt the search word of this reconstruct to search for, can improve the accuracy of search, effectively reduce user by continuous change search word or screening conditions be set the number of times that carries out repeat search, reduce the pressure of server.
Preferably, in order further to improve the accuracy of search, in above-mentioned steps S107, when server adopts the search word redefining to search for, can adopt the search word that redefines by step S102 selected go out commodity class search for now, also adopt the search word redefining to search for now in the commodity class of this user preference.
The searching method of the merchandise news providing for the embodiment of the present application above, based on same thinking, the embodiment of the present application also provides a kind of searcher of merchandise news, as shown in Figure 2.
The searcher structural representation of the merchandise news that Fig. 2 provides for the embodiment of the present application, specifically comprises:
Search classification determination module 201, for determining each commodity classification corresponding to search word of user's input;
Preference classification determination module 202 for the preference weights to each commodity classification according to predetermined described user, is selected commodity classification in each commodity classification of determining;
Divide word combination determination module 203, for the search word of described user's input is carried out to word segmentation processing, obtain each search participle, heading to the commodity class of selecting all commodity now carries out word segmentation processing, obtain each title participle, each search word based on obtaining and each title participle, build point word combination that includes at least one search participle and at least one title participle;
Standard words determination module 204, for according to point word combination of setting up in advance and the corresponding relation of standard words, determines respectively standard words corresponding to each point of word combination building, wherein, and the search word that standard words adopts while comprising each search in the past; For a standard words, the each participle having in point word combination of corresponding relation with this standard words is all participles of this standard words;
Set determination module 205, gathers to form a point set of words for the search participle and the title participle that respectively each point of word combination of the same standard words of correspondence are comprised;
Search word reconstructed module 206, redefines search word for the search participle and the title participle that comprise according to point set of words;
Described search classification determination module 201 also for, the search of at every turn carrying out for the past in advance, determine the search word adopting while carrying out this search, determine and adopt this search word to search for, and Search Results is returned to after the user of this search word of input, input the commodity classification under the merchandise news that the user of this search word clicks in Search Results, set up this commodity classification of determining and the corresponding relation of this search word;
Described search classification determination module 201 specifically for, according in advance set up search word and the corresponding relation of commodity classification, determine each commodity classification corresponding to search word that user inputs.
Described search classification determination module 201 also for, before setting up the corresponding relation of this definite commodity classification and this search word, the degree of relevancy value of this commodity classification and this search word is increased to default step value, and determine that the degree of relevancy value of current this commodity classification and this search word is not less than setting dependent thresholds.
Described preference classification determination module 202 specifically for, for each commodity classification, according to the assigned operation of described user past to the commodity execution now of this commodity class, determine the preference degree value of described user to this commodity classification, wherein, described assigned operation comprises: one or more in browse operation, order operation, search operation; To the preference degree value of each commodity classification order from big to small, in each commodity classification of determining, select to set successively the commodity classification of quantity according to described user.
Described search word reconstructed module 206 specifically for, from the each point of set of words of determining, filter out point set of words that comprises all search participles, and in the each point of set of words filtering out, select the maximum point set of words of quantity that comprises participle as optimal set, redefine search word according to the search participle comprising in described optimal set and title participle; Or, from the each point of set of words of determining, filter out point set of words that comprises all search participles, and for the each point of set of words filtering out, according to the weights of expressing the meaning of setting for each participle in advance, determine all participles that comprise in this point of set of words filtering out the weights of expressing the meaning and value, in the each point of set of words filtering out, select that determine with point set of words value maximum as optimal set, redefine search word according to the search participle comprising in described optimal set and title participle.
Described search module 207 specifically for, adopt the search word that redefines to search for now in the commodity class of selecting.
The searcher of concrete above-mentioned merchandise news can be arranged in server.
The embodiment of the present application provides a kind of searching method and device of merchandise news, the method is determined each commodity classification corresponding to search word that user inputs, therefrom select the commodity classification of user institute preference, respectively the search word of user's input and the heading of the commodity class of selecting all commodity are now carried out to participle, obtain each search participle and title participle, structure includes point word combination of at least one search participle and title participle, determine the each point of standard words that word combination is corresponding, determine in the each point of word combination by corresponding identical standard word and comprise point set of words that participle forms, and redefine search word according to the participle comprising in point set of words, the search word that employing redefines is searched for.The heading of the search word that said method is inputted based on user and the commodity class of user preference commodity now, search word to user's input has carried out reconstruct, the search word reconstructing had both comprised the search intention of this search of user, comprise again user's preference, therefore the search word based on reconstruct is searched for, can effectively improve the accuracy of the Search Results obtaining, thereby reduce the pressure of server.
Obviously, those skilled in the art can carry out various changes and modification and the spirit and scope that do not depart from the application to the application.Like this, if within these of the application are revised and modification belongs to the scope of the application's claim and equivalent technologies thereof, the application is also intended to comprise these changes and modification interior.
Claims (12)
1. a searching method for merchandise news, is characterized in that, comprising:
Determine each commodity classification corresponding to search word that user inputs;
Preference degree value according to predetermined described user to each commodity classification is selected commodity classification in each commodity classification of determining;
Search word to described user's input carries out word segmentation processing, obtain each search participle, heading to the commodity class of selecting all commodity now carries out word segmentation processing, obtain each title participle, each search participle based on obtaining and each title participle, build point word combination that includes at least one search participle and at least one title participle;
According to point word combination of setting up in advance and the corresponding relation of standard words, determine respectively standard words corresponding to each point of word combination building, wherein, the search word that standard words adopts while comprising each search in the past; For a standard words, the each participle having in point word combination of corresponding relation with this standard words is all participles of this standard words;
Respectively a point set of words gathered to form in the search participle and the title participle that in each point of word combination of the same standard words of correspondence, comprise;
Redefine search word according to the search participle comprising in point set of words and title participle;
The search word that employing redefines is searched for, and Search Results is returned to described user.
2. the method for claim 1, is characterized in that, before determining each commodity classification that search word that user inputs is corresponding, described method also comprises:
The search of at every turn carrying out for the past in advance, determine the search word adopting while carrying out this search, determine and adopt this search word to search for, and Search Results is returned to after the user of this search word of input, input the commodity classification under the merchandise news that the user of this search word clicks in Search Results, set up this commodity classification of determining and the corresponding relation of this search word;
Determine each commodity classification corresponding to search word that user inputs, specifically comprise:
According to the search word of setting up in advance and the corresponding relation of commodity classification, determine each commodity classification corresponding to search word that user inputs.
3. method as claimed in claim 2, is characterized in that, before setting up the corresponding relation of this definite commodity classification and this search word, described method also comprises:
The degree of relevancy value of this commodity classification and this search word is increased to default step value, and determine that the degree of relevancy value of current this commodity classification and this search word is not less than setting dependent thresholds.
4. the method for claim 1, is characterized in that, pre-determines the preference degree value of described user to each commodity classification, specifically comprises:
For each commodity classification, according to the assigned operation of described user past to the commodity execution now of this commodity class, determine the preference degree value of described user to this commodity classification, wherein, described assigned operation comprises: one or more in browse operation, order operation, search operation;
Preference degree value according to predetermined described user to each commodity classification is selected commodity classification in each commodity classification of determining, specifically comprises:
To the preference degree value of each commodity classification order from big to small, in each commodity classification of determining, select to set successively the commodity classification of quantity according to described user.
5. the method for claim 1, is characterized in that, redefines search word according to the search participle comprising in point set of words and title participle, specifically comprises:
From the each point of set of words of determining, filter out point set of words that comprises all search participles, and in the each point of set of words filtering out, select the maximum point set of words of quantity that comprises participle as optimal set, redefine search word according to the search participle comprising in described optimal set and title participle; Or
From the each point of set of words of determining, filter out point set of words that comprises all search participles, and for the each point of set of words filtering out, according to the weights of expressing the meaning of setting for each participle in advance, determine all participles that comprise in this point of set of words filtering out the weights of expressing the meaning and value, in the each point of set of words filtering out, select that determine with point set of words value maximum as optimal set, redefine search word according to the search participle comprising in described optimal set and title participle.
6. the method for claim 1, is characterized in that, adopts the search word redefining to search for, and specifically comprises:
The search word that employing redefines is searched for now in the commodity class of selecting.
7. a searcher for merchandise news, is characterized in that, comprising:
Search classification determination module, for determining each commodity classification corresponding to search word of user's input;
Preference classification determination module for the preference weights to each commodity classification according to predetermined described user, is selected commodity classification in each commodity classification of determining;
Divide word combination determination module, for the search word of described user's input is carried out to word segmentation processing, obtain each search participle, heading to the commodity class of selecting all commodity now carries out word segmentation processing, obtain each title participle, each search word based on obtaining and each title participle, build point word combination that includes at least one search participle and at least one title participle;
Standard words determination module, for according to point word combination of setting up in advance and the corresponding relation of standard words, determines respectively standard words corresponding to each point of word combination building, wherein, and the search word that standard words adopts while comprising each search in the past; For a standard words, the each participle having in point word combination of corresponding relation with this standard words is all participles of this standard words;
Set determination module, gathers to form a point set of words for the search participle and the title participle that respectively each point of word combination of the same standard words of correspondence are comprised;
Search word reconstructed module, redefines search word for the search participle and the title participle that comprise according to point set of words;
Search module, for adopting the search word redefining to search for, and returns to described user by Search Results.
8. device as claimed in claim 7, it is characterized in that, described search classification determination module also for, the search of at every turn carrying out for the past in advance, determine the search word adopting while carrying out this search, determine and adopt this search word to search for, and Search Results is returned to after the user of this search word of input, input the commodity classification under the merchandise news that the user of this search word clicks in Search Results, set up this commodity classification of determining and the corresponding relation of this search word;
Described search classification determination module specifically for, according in advance set up search word and the corresponding relation of commodity classification, determine each commodity classification corresponding to search word that user inputs.
9. device as claimed in claim 8, it is characterized in that, described search classification determination module also for, before setting up the corresponding relation of this definite commodity classification and this search word, the degree of relevancy value of this commodity classification and this search word is increased to default step value, and determine that the degree of relevancy value of current this commodity classification and this search word is not less than setting dependent thresholds.
10. device as claimed in claim 7, it is characterized in that, described preference classification determination module specifically for, for each commodity classification, according to the assigned operation of described user past to the commodity execution now of this commodity class, determine the preference degree value of described user to this commodity classification, wherein, described assigned operation comprises: one or more in browse operation, order operation, search operation; To the preference degree value of each commodity classification order from big to small, in each commodity classification of determining, select to set successively the commodity classification of quantity according to described user.
11. devices as claimed in claim 7, it is characterized in that, described search word reconstructed module specifically for, from the each point of set of words of determining, filter out point set of words that comprises all search participles, and in the each point of set of words filtering out, select the maximum point set of words of quantity that comprises participle as optimal set, redefine search word according to the search participle comprising in described optimal set and title participle; Or, from the each point of set of words of determining, filter out point set of words that comprises all search participles, and for the each point of set of words filtering out, according to the weights of expressing the meaning of setting for each participle in advance, determine all participles that comprise in this point of set of words filtering out the weights of expressing the meaning and value, in the each point of set of words filtering out, select that determine with point set of words value maximum as optimal set, redefine search word according to the search participle comprising in described optimal set and title participle.
12. devices as claimed in claim 7, is characterized in that, described search module specifically for, adopt the search word that redefines to search for now in the commodity class of selecting.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210448776.0A CN103810198B (en) | 2012-11-09 | 2012-11-09 | A kind of searching method of merchandise news and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210448776.0A CN103810198B (en) | 2012-11-09 | 2012-11-09 | A kind of searching method of merchandise news and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103810198A true CN103810198A (en) | 2014-05-21 |
| CN103810198B CN103810198B (en) | 2017-03-01 |
Family
ID=50706981
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210448776.0A Active CN103810198B (en) | 2012-11-09 | 2012-11-09 | A kind of searching method of merchandise news and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103810198B (en) |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104268265A (en) * | 2014-10-11 | 2015-01-07 | 时之我代信息科技（上海）有限公司 | Information searching method and information searching device |
| CN105159977A (en) * | 2015-08-27 | 2015-12-16 | 百度在线网络技术（北京）有限公司 | Information interaction processing method and apparatus |
| CN105488136A (en) * | 2015-11-25 | 2016-04-13 | 北京京东尚科信息技术有限公司 | Mining method of choosing hotspot tag |
| CN106547794A (en) * | 2015-09-22 | 2017-03-29 | 阿里巴巴集团控股有限公司 | Information search method and device |
| CN106682012A (en) * | 2015-11-06 | 2017-05-17 | 阿里巴巴集团控股有限公司 | Commodity object information searching method and device |
| CN106919576A (en) * | 2015-12-24 | 2017-07-04 | 北京奇虎科技有限公司 | Using the method and device of two grades of classes keywords database search for application now |
| CN106919587A (en) * | 2015-12-24 | 2017-07-04 | 北京奇虎科技有限公司 | Application program search system and method |
| CN107329964A (en) * | 2017-04-19 | 2017-11-07 | 阿里巴巴集团控股有限公司 | A kind of text handling method and device |
| CN108763579A (en) * | 2018-06-08 | 2018-11-06 | Oppo（重庆）智能科技有限公司 | Search for content recommendation method, device, terminal device and storage medium |
| CN108897785A (en) * | 2018-06-08 | 2018-11-27 | Oppo（重庆）智能科技有限公司 | Search for content recommendation method, device, terminal device and storage medium |
| CN109213921A (en) * | 2017-06-29 | 2019-01-15 | 广州涌智信息科技有限公司 | A kind of searching method and device of merchandise news |
| CN109522396A (en) * | 2018-10-22 | 2019-03-26 | 中国船舶工业综合技术经济研究院 | A kind of method of knowledge processing and system towards science and techniques of defence field |
| CN109543113A (en) * | 2018-12-21 | 2019-03-29 | 北京字节跳动网络技术有限公司 | Determine method, apparatus, storage medium and the electronic equipment clicked and recommend word |
| CN109635198A (en) * | 2018-12-17 | 2019-04-16 | 杭州柚子街信息科技有限公司 | The method, apparatus of presentation user's search result, medium and electronic equipment on merchandise display platform |
| CN109783733A (en) * | 2019-01-15 | 2019-05-21 | 三角兽（北京）科技有限公司 | User's portrait generating means and method, information processing unit and storage medium |
| CN110020157A (en) * | 2017-12-08 | 2019-07-16 | 北京京东尚科信息技术有限公司 | Data processing method, system, computer system and storage medium |
| CN110750985A (en) * | 2018-07-04 | 2020-02-04 | 阿里巴巴集团控股有限公司 | Brand word recognition method, device, equipment and storage medium |
| CN110781365A (en) * | 2018-07-13 | 2020-02-11 | 阿里巴巴集团控股有限公司 | Commodity searching method, device and system and electronic equipment |
| CN110795613A (en) * | 2018-07-17 | 2020-02-14 | 阿里巴巴集团控股有限公司 | Commodity searching method, device and system and electronic equipment |
| CN111159552A (en) * | 2019-12-30 | 2020-05-15 | 北京每日优鲜电子商务有限公司 | Commodity searching method, commodity searching device, server and storage medium |
| CN111400436A (en) * | 2020-02-19 | 2020-07-10 | 北京值得买科技股份有限公司 | Search method and device based on user intention recognition |
| CN113763123A (en) * | 2021-08-12 | 2021-12-07 | 阿里巴巴（中国）有限公司 | Commodity recommendation and search method, commodity recommendation and search equipment and storage medium |
| WO2023109292A1 (en) * | 2021-12-14 | 2023-06-22 | 北京字跳网络技术有限公司 | Search result display method and apparatus, computer device, and storage medium |
| CN117688147A (en) * | 2023-11-16 | 2024-03-12 | 深圳市灵智数字科技有限公司 | Commodity information searching method and device and terminal equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101206672A (en) * | 2007-12-25 | 2008-06-25 | 北京科文书业信息技术有限公司 | Commercial articles searching non result intelligent processing system and method |
| CN102446180A (en) * | 2010-10-09 | 2012-05-09 | 腾讯科技（深圳）有限公司 | Commodity searching method and device adopting same |
| US20120158691A1 (en) * | 2010-12-15 | 2012-06-21 | Electronics And Telecommunications Research Institute | Apparatus and method of searching hs codes using ontology |
-
2012
- 2012-11-09 CN CN201210448776.0A patent/CN103810198B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101206672A (en) * | 2007-12-25 | 2008-06-25 | 北京科文书业信息技术有限公司 | Commercial articles searching non result intelligent processing system and method |
| CN102446180A (en) * | 2010-10-09 | 2012-05-09 | 腾讯科技（深圳）有限公司 | Commodity searching method and device adopting same |
| US20120158691A1 (en) * | 2010-12-15 | 2012-06-21 | Electronics And Telecommunications Research Institute | Apparatus and method of searching hs codes using ontology |
Cited By (36)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104268265A (en) * | 2014-10-11 | 2015-01-07 | 时之我代信息科技（上海）有限公司 | Information searching method and information searching device |
| CN104268265B (en) * | 2014-10-11 | 2017-12-01 | 时之我代信息科技（上海）有限公司 | A kind of information search method and device |
| CN105159977B (en) * | 2015-08-27 | 2019-01-25 | 百度在线网络技术（北京）有限公司 | Information interactive processing method and device |
| CN105159977A (en) * | 2015-08-27 | 2015-12-16 | 百度在线网络技术（北京）有限公司 | Information interaction processing method and apparatus |
| CN106547794A (en) * | 2015-09-22 | 2017-03-29 | 阿里巴巴集团控股有限公司 | Information search method and device |
| CN106547794B (en) * | 2015-09-22 | 2020-04-14 | 阿里巴巴集团控股有限公司 | Information searching method and device |
| CN106682012A (en) * | 2015-11-06 | 2017-05-17 | 阿里巴巴集团控股有限公司 | Commodity object information searching method and device |
| CN105488136A (en) * | 2015-11-25 | 2016-04-13 | 北京京东尚科信息技术有限公司 | Mining method of choosing hotspot tag |
| CN105488136B (en) * | 2015-11-25 | 2019-03-26 | 北京京东尚科信息技术有限公司 | Choose the method for digging of hot spot label |
| CN106919587A (en) * | 2015-12-24 | 2017-07-04 | 北京奇虎科技有限公司 | Application program search system and method |
| CN106919587B (en) * | 2015-12-24 | 2021-01-29 | 北京奇虎科技有限公司 | Application program searching system and method |
| CN106919576A (en) * | 2015-12-24 | 2017-07-04 | 北京奇虎科技有限公司 | Using the method and device of two grades of classes keywords database search for application now |
| CN107329964B (en) * | 2017-04-19 | 2021-01-05 | 创新先进技术有限公司 | Text processing method and device |
| CN107329964A (en) * | 2017-04-19 | 2017-11-07 | 阿里巴巴集团控股有限公司 | A kind of text handling method and device |
| CN109213921A (en) * | 2017-06-29 | 2019-01-15 | 广州涌智信息科技有限公司 | A kind of searching method and device of merchandise news |
| CN110020157A (en) * | 2017-12-08 | 2019-07-16 | 北京京东尚科信息技术有限公司 | Data processing method, system, computer system and storage medium |
| CN108763579B (en) * | 2018-06-08 | 2020-12-22 | Oppo（重庆）智能科技有限公司 | Search content recommendation method and device, terminal device and storage medium |
| CN108763579A (en) * | 2018-06-08 | 2018-11-06 | Oppo（重庆）智能科技有限公司 | Search for content recommendation method, device, terminal device and storage medium |
| CN108897785A (en) * | 2018-06-08 | 2018-11-27 | Oppo（重庆）智能科技有限公司 | Search for content recommendation method, device, terminal device and storage medium |
| CN110750985B (en) * | 2018-07-04 | 2023-07-11 | 阿里巴巴集团控股有限公司 | Brand word recognition method, device, equipment and storage medium |
| CN110750985A (en) * | 2018-07-04 | 2020-02-04 | 阿里巴巴集团控股有限公司 | Brand word recognition method, device, equipment and storage medium |
| CN110781365B (en) * | 2018-07-13 | 2023-04-28 | 阿里巴巴集团控股有限公司 | Commodity searching method, device and system and electronic equipment |
| CN110781365A (en) * | 2018-07-13 | 2020-02-11 | 阿里巴巴集团控股有限公司 | Commodity searching method, device and system and electronic equipment |
| CN110795613A (en) * | 2018-07-17 | 2020-02-14 | 阿里巴巴集团控股有限公司 | Commodity searching method, device and system and electronic equipment |
| CN110795613B (en) * | 2018-07-17 | 2023-04-28 | 阿里巴巴集团控股有限公司 | Commodity searching method, device and system and electronic equipment |
| CN109522396A (en) * | 2018-10-22 | 2019-03-26 | 中国船舶工业综合技术经济研究院 | A kind of method of knowledge processing and system towards science and techniques of defence field |
| CN109635198B (en) * | 2018-12-17 | 2020-09-29 | 杭州柚子街信息科技有限公司 | Method, device, medium and electronic equipment for presenting user search results on commodity display platform |
| CN109635198A (en) * | 2018-12-17 | 2019-04-16 | 杭州柚子街信息科技有限公司 | The method, apparatus of presentation user's search result, medium and electronic equipment on merchandise display platform |
| CN109543113B (en) * | 2018-12-21 | 2022-02-01 | 北京字节跳动网络技术有限公司 | Method and device for determining click recommendation words, storage medium and electronic equipment |
| CN109543113A (en) * | 2018-12-21 | 2019-03-29 | 北京字节跳动网络技术有限公司 | Determine method, apparatus, storage medium and the electronic equipment clicked and recommend word |
| CN109783733A (en) * | 2019-01-15 | 2019-05-21 | 三角兽（北京）科技有限公司 | User's portrait generating means and method, information processing unit and storage medium |
| CN111159552A (en) * | 2019-12-30 | 2020-05-15 | 北京每日优鲜电子商务有限公司 | Commodity searching method, commodity searching device, server and storage medium |
| CN111400436A (en) * | 2020-02-19 | 2020-07-10 | 北京值得买科技股份有限公司 | Search method and device based on user intention recognition |
| CN113763123A (en) * | 2021-08-12 | 2021-12-07 | 阿里巴巴（中国）有限公司 | Commodity recommendation and search method, commodity recommendation and search equipment and storage medium |
| WO2023109292A1 (en) * | 2021-12-14 | 2023-06-22 | 北京字跳网络技术有限公司 | Search result display method and apparatus, computer device, and storage medium |
| CN117688147A (en) * | 2023-11-16 | 2024-03-12 | 深圳市灵智数字科技有限公司 | Commodity information searching method and device and terminal equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103810198B (en) | 2017-03-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103810198A (en) | Method and device for searching commodity information | |
| CN102193936B (en) | Data classification method and device | |
| JP6247292B2 (en) | Query expansion | |
| CN107180093A (en) | Information search method and device and ageing inquiry word recognition method and device | |
| CN106250513A (en) | A kind of event personalization sorting technique based on event modeling and system | |
| CN103942712A (en) | Product similarity based e-commerce recommendation system and method thereof | |
| CN103729359A (en) | Method and system for recommending search terms | |
| CN103578008A (en) | Method and device for recommending garment products | |
| CN106600372A (en) | Commodity recommending method and system based on user behaviors | |
| CN106547864B (en) | A kind of Personalized search based on query expansion | |
| CN105302807B (en) | Method and device for acquiring information category | |
| CN105956148A (en) | Resource information recommendation method and apparatus | |
| CN103530339A (en) | Mobile application information push method and device | |
| CN104252456A (en) | Method, device and system for weight estimation | |
| CN105023178B (en) | A kind of electronic commerce recommending method based on ontology | |
| CN104778237A (en) | Individual recommending method and system based on key users | |
| CN104077417A (en) | Figure tag recommendation method and system in social network | |
| CN103870575A (en) | Method and device for extracting domain keywords | |
| CN107277115A (en) | A kind of content delivery method and device | |
| CN103136213A (en) | Method and device for providing related words | |
| CN109522396B (en) | Knowledge processing method and system for national defense science and technology field | |
| CN108427756A (en) | Personalized query word completion recommendation method and device based on same-class user model | |
| CN105740480A (en) | Air ticket recommending method and system | |
| CN103336765B (en) | A kind of markov matrix off-line correction method of text key word | |
| AU2021385976A1 (en) | Product recommendation method and apparatus, computer storage medium, 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 | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 1195374 Country of ref document: HK |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: GR Ref document number: 1195374 Country of ref document: HK |