CN102799589A - Information pushing method and device - Google Patents
Information pushing method and device Download PDFInfo
- Publication number
- CN102799589A CN102799589A CN201110138313XA CN201110138313A CN102799589A CN 102799589 A CN102799589 A CN 102799589A CN 201110138313X A CN201110138313X A CN 201110138313XA CN 201110138313 A CN201110138313 A CN 201110138313A CN 102799589 A CN102799589 A CN 102799589A
- Authority
- CN
- China
- Prior art keywords
- user
- matching degree
- keyword
- weight
- news
- 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
Abstract
The invention discloses an information pushing method and device. The method comprises the following steps of: extracting a keyword from a problem submitted by a user; obtaining publishing information of the keyword from historical behavior data of the user according to a text matching method; counting the relative user related to the publishing information from the historical behavior data of the user according to a counting and analyzing method; calculating a matching degree between the problem and the relative user; and pushing the problem to the user with the initial preconfigured quantity of the relative user with the highest matching degree. According to the embodiment of the invention, the feedback speed of a knowledge encyclopedia system to answers can be improved and the accuracy of answering is improved.
Description
Technical field
The application relates to communication and field of computer technology, particularly relates to a kind of information-pushing method and device.
Background technology
The various problem that the user proposes can be answered by existing knowledge encyclopaedia system; If knowledge encyclopaedia system platform receives the up-to-date issue message that a client is sent; Knowledge encyclopaedia system extracts several keywords from the up-to-date issue message that client is submitted to; From database, search the historical problem message that comprises at least one keyword that extracts; Further search the client of these historical problem message of issue again, giving tacit consent to these clients is exactly to answer the client of this up-to-date issue message, at last this up-to-date issue message is sent to all clients that find.
But; The inventor finds under study for action; All historical problem message that comprise the keyword that extracts need be searched by existing knowledge encyclopaedia system when searching the client that can answer up-to-date problem, and further search all clients of these historical problem message of issue.Often, knowledge encyclopaedia system searching to the quantity of client be huge, therefore; In search procedure; Knowledge encyclopaedia system need consume a large amount of resource of server, has not only increased load of server, and the processing power of server is also had very high requirement.And, if knowledge encyclopaedia system is distributed to all clients that find with this up-to-date issue message, the data volume of transmission through network is increased, be prone to cause network blockage, finally increased the burden of network.
Summary of the invention
In order to solve the problems of the technologies described above, the application embodiment provides a kind of information-pushing method and device, to reduce server and the network burden of knowledge encyclopaedia system in the process of answering a question.
The openly following technical scheme of the application embodiment:
A kind of information-pushing method comprises:
From the problem that the user submits to, extract keyword;
From user's historical behavior data, obtain releasing news of said keyword according to the text matches method;
From user's historical behavior data, add up and the said relevant associated user that releases news according to statistical analysis technique;
Calculate the matching degree between said problem and the said associated user, said problem is pushed to the user of the preset number that begins from the highest associated user of matching degree.
A kind of information push-delivery apparatus comprises:
Extraction unit is used for extracting keyword from the problem that the user submits to;
Acquiring unit is used for obtaining releasing news of said keyword according to the text matches method from user's historical behavior data;
Statistic unit is used for according to the historical behavior data statistics and said release news relevant associated user of statistical analysis technique from the user;
Push unit is used to calculate the matching degree between said problem and the said associated user, said problem is pushed to the user of the preset number that begins from the highest associated user of matching degree.
Can find out that by the foregoing description after new problem was submitted to, the associated user relevant with releasing news of problem can search in system, and then be pushed to problem in face of these associated users targetedly.For server, server only pushes " problem " to the specific user, rather than all pushes problem to all users, therefore, has reduced the consumption of server resource, has reduced load of server.In addition,, also just reduced the data conveying capacity in the network, improved network transfer speeds, reduced the burden of network owing to only push problem to particular user.
And, to compare single passive text matches method and solve efficient in problem, the popularity and the accuracy of answer viewpoint all improve a lot, and the solution of commercial knowledge encyclopaedia problem and the popularization of answer are had more directiveness.
Description of drawings
In order to be illustrated more clearly in the application embodiment or technical scheme of the prior art; To do to introduce simply to the accompanying drawing of required use in embodiment or the description of the Prior Art below; Obviously; For those of ordinary skills, under the prerequisite of not paying creative work property, can also obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is the process flow diagram of an embodiment of a kind of information-pushing method of the application;
Fig. 2 is for being pushed to problem according to matching degree the process flow diagram of user's a embodiment among the application;
Fig. 3 is for being pushed to problem according to matching degree the process flow diagram of another embodiment of user among the application;
Fig. 4 is for being pushed to problem according to matching degree the process flow diagram of another embodiment of user among the application;
Fig. 5 is the process flow diagram of another embodiment of a kind of information-pushing method of the application;
Fig. 6 is the structural drawing of an enforcement of a kind of information push-delivery apparatus of the application;
Fig. 7 is a structural drawing of push unit among the application;
Fig. 8 is another structural drawing of push unit among the application;
Fig. 9 is another structural drawing of push unit among the application.
Embodiment
For above-mentioned purpose, the feature and advantage that make the application can be more obviously understandable, the application embodiment is described in detail below in conjunction with accompanying drawing.
Embodiment one
See also Fig. 1, it is the process flow diagram of an embodiment of a kind of information-pushing method of the application, and this method may further comprise the steps:
Step 101: from the problem that the user submits to, extract keyword;
At user's page that submits a question, the user is except submitting a question, and system also can require the user in the input problem content, and the classification under the given problem is set up the corresponding relation of problem and classification.For example, can in the interface that submits a question, set up the classification drop-down list, supply the user in drop-down list, to specify the affiliated classification of the problem of being submitted to.
In addition; After the user submits a question success; In order to guarantee the accuracy of the classification that the user submits to, auditing flow can also further be got into, by the correct top of operation personnel's manual examination and verification problem and classification corresponding relation; If the corresponding relation mistake can be revised the corresponding relation of problem and classification by the operation personnel.
The problem that the user submits to is submitted to after the webserver, by the semanteme of the webserver according to problem, the keyword in the extraction problem.
Need to prove, in the application's technical scheme, the quantity of the keyword that extracts is not limited, can set the quantity of the keyword that extracts arbitrarily according to the needs in the different system.In addition, can adopt any method that is implemented in of the prior art according to the extraction of semantics keyword, the application does not do concrete qualification to method for distilling yet yet.
Step 102: from user's historical behavior data, obtain release news relevant with said keyword according to the text matches method;
Releasing news is meant the production of user's commercial product of issue in the website or wants to buy information, is distinguished by the sign (offer_ID) that releases news between each releases news.In addition, from user's historical behavior data, can also get access to each affiliated classification that releases news.
In order to reduce information processing capacity, except can obtaining whole release news relevant, preferred with each keyword; For each keyword; Can sort to releasing news according to keyword and the matching degree that releases news, according to matching degree order from high to low, from the maximum beginning that releases news of matching degree; According to the needs in the different system, obtain releasing news of some.
Need to prove; In the technical scheme of application; Can adopt any text matches method that is implemented in the existing search technique to obtain release news relevant with each keyword, the application does not limit the method for distilling itself that releases news.
Step 103: from user's historical behavior data, add up and the said relevant associated user that releases news according to statistical analysis technique;
Wherein, can be this user who releases news of issue with the relevant associated user that releases news, also can be this user who releases news of feedback, can also be to browse or inquire about this user who releases news etc.According to statistical analysis technique, in each user's who from database, is write down the historical behavior data, count the relevant associated user that releases news who obtains with each.
Need to prove; In the application's technical scheme; Can adopt any statistical analysis technique that is implemented in the existing search technique to obtain and each relevant associated user that releases news, the application does not limit acquisition methods the application of associated user.
Step 104: calculate the matching degree between said problem and the said associated user, said problem is pushed to the user of the preset number that begins from the highest associated user of matching degree.
In all associated users that obtain with each releases news relevant, calculate this problem and the matching degree between each associated user that the user submits to, at last, problem is pushed to from the highest associated user of matching degree begins the user of preset number.For example, through statistics, have 100 with each relevant associated user that releases news; Calculate problem that the user submits to and these 100 associated users' matching degree respectively, 100 associated users' matching degree is sorted according to from high to low order, the associated user the highest from matching degree begins; Extract the associated user of preset number; As extract 10 associated users, last, this problem that the user is submitted to is pushed to this 10 users.
Need to prove, in the technical scheme of application, the associated user's that extracts quantity is not limited, can set the associated user's who extracts quantity arbitrarily according to the needs in the different system.
Preferably; When being this user who releases news of issue with the relevant associated user that releases news; Matching degree between said problem of said calculating and the said associated user; Said problem is pushed to the highest associated user of matching degree associated user that begin, preset number may further comprise the steps, as shown in Figure 2
S1041A: according to keyword importance keyword assignment weight for extracting to problem in the semanteme statement, wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big;
For example, list the corresponding relation between importance and the weight with the form of tabulation, wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big.For some keywords, when obtaining this keyword after the importance to problem on the semantic meaning representation, the corresponding weight of importance of searching and obtaining through the corresponding relation tabulation.
S1042A: the matching degree according to releasing news with keyword is that releasing news of obtaining assigns weight, and wherein, the weight that release news lower than matching degree that release news that matching degree is high is big;
For example, list the corresponding relation between the matching degree weight with the form of tabulation, wherein, the weight that release news lower than matching degree that release news that matching degree is high is big.For some releasing news, after obtaining this matching degree with keyword of releasing news, the corresponding weight of matching degree of searching and obtaining through the corresponding relation tabulation.
S1043A: the product of the weight of calculating keyword and the weight that releases news of said keyword; Again with matching degree and the summation of said product value between the classification of problem and the classification that releases news, with said result as problem and first matching degree of issuing between the said user who releases news;
S1044A: the order according to from high to low sorts to said first matching degree, and problem is sent to since the user of the highest user's of first matching degree preset number.
Suppose that the keyword quantity that from the problem that the user submits to, extracts is n; Be respectively Kw1, Kw2, Kw3......Kwn; Be respectively n keyword assignment weight according to n keyword importance to problem in the semanteme statement; The weight of the keyword that the keyword that importance is high is lower than importance is big, and therefore, the weight of the keyword assignment that importance is the highest is maximum; The weight of the keyword assignment that importance is minimum is minimum, and importance equates or close keyword can distribute identical weighted value.
At this, it is emphasized that in the application's technical scheme the concrete numerical value to the weight of distributing to each keyword does not limit, as long as the weight that satisfies the high keyword of the importance keyword lower than importance greatly.
For example, if be that keyword sorts, can be primary keyword assignment weighted value 0.5 according to importance order from high to low; Be deputy keyword assignment weighted value 0.3, be tertiary keyword assignment weighted value 0.2, when the importance of remaining keyword differs very little; When thinking close; Can be for remaining keyword assignment weighted value 0.1, as shown in the table, following table is the allocation result of weights.
| Kw1 | Kw2 | Kw3 | Kw4 | ?..... | Kwn |
| 0.5 | 0.3 | 0.2 | 0.1 | 0.1 | 0.1 |
Suppose for keyword Kw1; The quantity of obtaining that releases news is m; Being respectively offer1, offer2, offer3......offerm, is that releasing news of obtaining assigns weight according to the matching degree that releases news with keyword, and the weight that release news lower than matching degree that release news that matching degree is high is big; Therefore; The weight maximum that releases news and distribute that matching degree is the highest, the weight minimum that releases news and distribute that matching degree is minimum, matching degree equates or close releasing news can distribute identical weighted value.
At this, also it is emphasized that in the application's technical scheme, the concrete numerical value of distributing to each weight that releases news is not limited, as long as satisfy high the releasing news greatly of matching degree than the low weight that releases news of matching degree.
For example, as shown in the table, following table is the allocation result that releases news.
| ?offer?1 | offer?2 | offer?3 | offer?4 | offer?5 | ..... | offer?m |
| 1.5 | 1.4 | 1.3 | 1.2 | 1.1 | 1.0 | 1.0 |
After having calculated weight; According to first matching degree between formula Match_Offer_Owner=Weight_KW * problem that Weight_Offer+Category_match calculating user submits to and the user who releases news; Wherein, Weight_KW is the weight of keyword, and Weight_Offer is the weight that releases news, matching degree between the classification of the problem that Category_match submits to for the user and the classification that releases news.If the classification of the problem that the user submits to is identical with the classification that releases news; The classification of the problem that the user submits to and the classification that releases news coupling; Matching degree is 1; If the classification of the problem that the user submits to is different with the classification that releases news, the classification of the problem that the user submits to does not match with the classification that releases news, and matching degree is 0.
After first matching degree between the user who releases news of the problem that calculates user's submission and this problem of issue; Order according to from high to low sorts to first matching degree; Problem is sent to since the highest user of first matching degree user of preset number.
Another kind of preferred mode is; If relevant associated user is this user who releases news of feedback with releasing news; Matching degree between said problem of said calculating and the said associated user; Said problem is pushed to the highest associated user of matching degree associated user that begin, preset number comprises the steps, of Fig. 3
S1041B: according to keyword importance keyword assignment weight for extracting to problem in the semanteme statement, wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big;
S1042B: the matching degree according to releasing news with keyword is that releasing news of obtaining assigns weight, and wherein, the weight that release news lower than matching degree that release news that matching degree is high is big;
S1043B: from user's historical behavior data, add up the said user who releases news of feedback to the said degree of feedback that releases news according to statistical analysis technique;
For example; Be the production information of certain commercial product if release news; Then user feedback releases news and just is meant that user's expression that releases news need to buy this commercial product; Be the information of wanting to buy of certain commercial product if release news, then user feedback releases news and just is meant that user's expression that releases news need to sell this commercial product.
S1044B: the weight, the weight that releases news of said keyword and the product of the said degree of feedback that calculate keyword; Again with matching degree and the summation of said product value between the classification of problem and the classification that releases news, with said result as said problem and second matching degree of feeding back between the said user who releases news;
S1045B: the order according to from high to low will sort to said second matching degree, and problem is sent to since the highest user user, preset number of first matching degree.
Wherein, The user that feedback releases news has been reflected the attention rate of this user to releasing news to the degree of feedback that releases news; Preferably; A kind of method of calculating the degree of feedback is: the product of the said user's who releases news of calculating feedback the feedback number of times and the ageing factor of said feedback, wherein, the ageing factor of the feedback that the feedback ratio distance current timing statistics near apart from current timing statistics is far away is big.
For example; User A carried out feedback 10 times to offer1 before 30 days; And user B has carried out feedback at nearest 3 days 10 times to offer1; Because the current timing statistics of the feedback of user B distance is nearer than the current timing statistics of feedback distance of user A, therefore, the ageing factor of the feedback of user B is bigger than the ageing factor of the feedback of user A.As, the ageing factor of setting the feedback of user B is 0.5, and the ageing factor of the feedback of user A is 0.3, and then being used for B is 0.5 * 10=5 to the degree of feedback of offer1, and the degree of feedback of user A is 0.3 * 10=3.
It is emphasized that in the application's technical scheme the concrete numerical value to the ageing factor of each feedback user does not limit, as long as satisfy the near current timing statistics of feedback ratio distance of the current timing statistics of the distance ageing factor far away greatly.
After having calculated the weight and the degree of feedback; Calculate the problem of user's submission and feed back second matching degree between the user who releases news according to formula Match_Offer_Owner=Weight_KW * Weight_Offer * Feedback+Category_match; Wherein, Weight_KW is the weight of keyword; Weight_Offer is the weight that releases news, and Feedback is the degree of feedback of user to releasing news that release news of feedback, and Category_match is matching degree between the classification of the problem of user's submission and the classification that releases news.If the classification of the problem that the user submits to is identical with the classification that releases news; The classification of the problem that the user submits to and the classification that releases news coupling; Matching degree is 1; If the classification of the problem that the user submits to is different with the classification that releases news, the classification of the problem that the user submits to does not match with the classification that releases news, and matching degree is 0.
After second matching degree between the user who releases news of the problem that calculates user's submission and this problem of feedback; Order according to from high to low sorts to second matching degree; Problem is sent to since the highest user of second matching degree user of preset number.
Also have a kind of preferred mode to be; If both comprised this user who releases news of issue with the relevant associated user that releases news; Also comprise this user who releases news of feedback, the matching degree between said problem of said calculating and the said associated user is pushed to the highest associated user of matching degree associated user that begin, preset number with said problem and may further comprise the steps; As shown in Figure 4
S1041C: according to keyword importance keyword assignment weight for extracting to problem in the semanteme statement, wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big;
S1042C: the matching degree according to releasing news with keyword is that releasing news of obtaining assigns weight, and wherein, the weight that release news lower than matching degree that release news that matching degree is high is big;
S1043C: from user's historical behavior data, add up the said user who releases news of feedback to the said degree of feedback that releases news according to statistical analysis technique;
S1044C: the product of the weight of calculating keyword and the weight that releases news of said keyword; Again with matching degree and the summation of said product value between the classification of problem and the classification that releases news; Obtain first matching degree between said problem and the said user who releases news of feedback; And; Calculate weight, the weight that releases news of said keyword and the product of the said degree of feedback of keyword, with matching degree and the summation of said product value between the classification of problem and the classification that releases news, obtain said problem and issue second matching degree between the said user who releases news again;
S1045C: the order according to from high to low sorts to said first matching degree and second matching degree, and problem is sent to the user that begin, preset number from the highest user of matching degree.
If both comprised this user who releases news of issue with the relevant associated user that releases news; Also comprise this user who releases news of feedback; In step S1045C, need unite ordering to first matching degree and second matching degree; At last, problem is sent to from the highest user of matching degree begin the user of preset number.
Can find out that by the foregoing description after new problem was submitted to, the associated user relevant with releasing news of problem can search in system, and then be pushed to problem in face of these associated users targetedly.For server, server only pushes " problem " to the specific user, rather than all pushes problem to all users, therefore, has reduced the consumption of server resource, has reduced load of server.In addition,, also just reduced the data conveying capacity in the network, improved network transfer speeds, reduced the burden of network owing to only push problem to particular user.
Release news and, compare single passive text matches method and solve efficient in problem, the popularity and the accuracy of answer viewpoint all improve a lot, and the solution of commercial knowledge encyclopaedia problem and the popularization of answer are had more directiveness.
Embodiment two
Be the whole process that example specifies information push below with the special scenes.See also Fig. 5, it is the process flow diagram of another embodiment of a kind of information-pushing method of the application, may further comprise the steps:
Step 501: the user submits a problem to web page server, and in question marks purpose drop-down list, specifies an affiliated classification for this problem;
Step 502: web page server extracts keyword from the problem of submitting to, according to keyword importance keyword assignment weight for extracting to this problem in the semanteme statement;
For example; After extracting keyword, web page server can sort to keyword according to each keyword importance degree to this problem in the semanteme statement; The keyword that importance is the highest makes number one, and the keyword that importance is minimum rolls into last place.Then, web page server is each keyword assignment weight according to ordering, and wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big.
In the numerous keywords that from certain problem, extract, in fact, the higher keyword of only making a difference has the value of use.For fear of the treatment effeciency problem of when handling a large amount of keywords, being brought; A kind ofly preferred embodiment be; Be each keyword assignment after the weight, can beginning that the keyword that only extracts some carries out follow-up processing from the maximum keyword of weight.Extracting quantity can set arbitrarily according to the needs of system.Like this, can improve the treatment effeciency of information push.
Step 503: web page server calls the text matches interface successively, and the text matches interface obtains releasing news of each keyword according to the text matches method from user's historical behavior data;
In addition, in step 502, when web page server had only extracted the keyword of some from all keywords, in this step, the text matches interface also only obtained releasing news of this part key speech.
In addition, the text matches interface except obtain release news, also from user's historical behavior data, obtain the classification under releasing news.
Above-mentioned text matches interface can be the text matches interface that is provided in the existing search technique.
Step 504: the text matches interface is that releasing news of obtaining assigns weight according to keyword importance to problem in the semanteme statement;
For example; The text matches interface can sort to releasing news of each keyword according to keyword and the matching degree that releases news; Assign weight for each releases news according to ranking results, the weight that release news lower than matching degree that release news that matching degree is high is big.
As, for the simple assigning process of describing weight, suppose that for keyword Kw1, the text matches interface has obtained 6 and released news: offer1-offer6.According to Kw1 and 6 matching degrees that release news offer1-offer6 is sorted, releasing news that matching degree is the highest makes number one, and releasing news that matching degree is minimum rolls into last place.Then, the text matches interface is followed successively by 6 according to the result of ordering and releases news and assign weight, and the weight that release news lower than matching degree that release news that matching degree is high is big.
Certainly; In practical application; Releasing news of each keyword not only has only 6, when some keywords have a large amount of releasing news, and can be from the maximum beginning that releases news of weight; Only extract releasing news of some, only assign weight and carry out follow-up processing for this part that extract releases news.Extracting quantity can set arbitrarily according to the needs of system.Like this, can improve the treatment effeciency of information push.
Step 505: web page server calls the inquiring user module, from user's historical behavior data, is added up issue or feeds back the user that each releases news according to statistical analysis technique by inquiring user module module;
In step 504, when the text matches module was only extracted releasing news of some, inquiring user module module was only added up issue or is fed back the user that this part releases news.
Above-mentioned inquiring user module module can be the inquiring user module module that is provided in the existing search technique.
Step 506: inquiring user module module is added up the degree of feedback of user to releasing news that feedback releases news according to statistical analysis technique from user's historical behavior data;
In step 505, when inquiring user module module is only added up the user that a feedback part releases news, the degree of feedback of user that release news of this part of statistics feedback then to releasing news.
Step 507: web page server calculates first matching degree between the user that problem that the user submits to and this problem of issue release news respectively, and, second matching degree between the user that the problem that the user submits to and this problem of feedback release news;
Wherein, The product of the weight of calculating keyword and the weight that releases news of said keyword; With matching degree and the summation of said product value between the classification of problem and the classification that releases news, obtain said problem and first matching degree of issuing between the said user who releases news again.
Calculate weight, the weight that releases news of said keyword and the product of the said degree of feedback of keyword; With matching degree and the summation of said product value between the classification of problem and the classification that releases news, obtain said problem and second matching degree of feeding back between the said user who releases news again.
Step 508: web page server sorts to first matching degree and second matching degree that calculates according to order from high to low, problem is sent to the user of the preset number that begins from the highest user of matching degree.
Pushing number of users sets arbitrarily with the needs according to system.
Can find out that by the foregoing description after new problem was submitted to, the associated user relevant with releasing news of problem can search in system, and then be pushed to problem in face of these associated users targetedly.For server, server only pushes " problem " to the specific user, rather than all pushes problem to all users, therefore, has reduced the consumption of server resource, has reduced load of server.In addition,, also just reduced the data conveying capacity in the network, improved network transfer speeds, reduced the burden of network owing to only push problem to particular user.
Release news
And, to compare single passive text matches method and solve efficient in problem, the popularity and the accuracy of answer viewpoint all improve a lot, and the solution of commercial knowledge encyclopaedia problem and the popularization of answer are had more directiveness.
Embodiment three
Corresponding with above-mentioned a kind of information-pushing method, the application embodiment also provides a kind of information push-delivery apparatus.See also Fig. 6, it is the structural drawing of an embodiment of a kind of information push-delivery apparatus of the application, and this device comprises extraction unit 601, acquiring unit 602, statistic unit 603 and push unit 604.Principle of work below in conjunction with this device is further introduced its inner structure and annexation.
Extraction unit 601 is used for extracting keyword from the problem that the user submits to;
Acquiring unit 602 is used for obtaining releasing news of said keyword according to the text matches method from user's historical behavior data;
Statistic unit 603 is used for according to the historical behavior data statistics and said release news relevant associated user of statistical analysis technique from the user;
Push unit 604 is used to calculate the matching degree between said problem and the said associated user, said problem is pushed to the user of the preset number that begins from the highest associated user of matching degree.
Preferably, see also Fig. 7, it is a structural drawing of push unit among the application; As shown in Figure 7; Push unit 604 comprises: the first weight allocation subelement 6041, second weight allocation in the unit 6042, the first matching degree computation subunit 6043 and first send subelement 6044, wherein
The first weight allocation subelement 6041 is used for according to keyword importance keyword assignment weight for extracting to problem in the semanteme statement, and wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big;
The second weight allocation subelement 6042, being used for according to the matching degree that releases news with keyword is that releasing news of obtaining assigns weight, wherein, the weight that release news lower than matching degree that release news that matching degree is high is big;
The first matching degree computation subunit 6043; Be used to calculate the product of the weight that releases news of weight and the said keyword of keyword; With matching degree and the summation of said product value between the classification of problem and the classification that releases news, obtain said problem and first matching degree of issuing between the said user who releases news again;
First sends subelement 6044, is used for according to from high to low order said first matching degree being sorted, and problem is sent to since the highest user user, preset number of first matching degree.
Except structure shown in Figure 7; Preferably, see also Fig. 8, it is another structural drawing of push unit among the application; As shown in Figure 8; Push unit 604 comprises: the first weight allocation subelement 6041, second weight allocation in the unit 6042, degree of feedback statistics subelement 6045, the second matching degree computation subunit 6046 and second send subelement 6047, wherein
The first weight allocation subelement 6041 is used for according to keyword importance keyword assignment weight for extracting to problem in the semanteme statement, and wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big;
The second weight allocation subelement 6042, being used for according to the matching degree that releases news with keyword is that releasing news of obtaining assigns weight, wherein, the weight that release news lower than matching degree that release news that matching degree is high is big;
Degree of feedback statistics subelement 6045 is used for feeding back the said user who releases news to the said degree of feedback that releases news according to statistical analysis technique from user's historical behavior data statistics;
The second matching degree computation subunit 6046; Be used to calculate the weight that releases news of the weight of keyword, said keyword and the product of the said degree of feedback; With matching degree and the summation of said product value between the classification of problem and the classification that releases news, obtain said problem and second matching degree of feeding back between the said user who releases news again;
Second sends subelement 6047, is used for will sorting to said second matching degree according to from high to low order, and problem is sent to since the user of the highest user's of second matching degree preset number.
Except Fig. 7 and structure shown in Figure 8; Further preferred, see also Fig. 9, it is another structural drawing of push unit among the application; As shown in Figure 9; Push unit 604 comprises: the first weight allocation subelement 6041, second weight allocation in the unit 6042, degree of feedback statistics subelement 6045, COMPREHENSIVE CALCULATING subelement 6048 and the 3rd send subelement 6049, wherein
The first weight allocation subelement 6041 is used for according to keyword importance keyword assignment weight for extracting to problem in the semanteme statement, and wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big;
The second weight allocation subelement 6042, being used for according to the matching degree that releases news with keyword is that releasing news of obtaining assigns weight, wherein, the weight that release news lower than matching degree that release news that matching degree is high is big;
Degree of feedback statistics subelement 6045 is used for feeding back the said user who releases news to the said degree of feedback that releases news according to statistical analysis technique from user's historical behavior data statistics;
The 3rd sends subelement 6049, is used for according to from high to low order said first matching degree and second matching degree being sorted, and problem is sent to the user of the preset number that begins from the highest user of matching degree.
Further preferred; In Fig. 8 and push unit shown in Figure 9; Degree of feedback statistics subelement comprises: computation subunit; Be used to calculate the said user's who releases news of feedback the feedback number of times and the product of the ageing factor of feedback, wherein, the ageing factor of the feedback that the near feedback ratio of the current timing statistics of distance is far away apart from current timing statistics is big.
Can find out that by the foregoing description after new problem was submitted to, the associated user relevant with releasing news of problem can search in system, and then be pushed to problem in face of these associated users targetedly.For server, server only pushes " problem " to the specific user, rather than all pushes problem to all users, therefore, has reduced the consumption of server resource, has reduced load of server.In addition,, also just reduced the data conveying capacity in the network, improved network transfer speeds, reduced the burden of network owing to only push problem to particular user.
Release news and, compare single passive text matches method and solve efficient in problem, the popularity and the accuracy of answer viewpoint all improve a lot, and the solution of commercial knowledge encyclopaedia problem and the popularization of answer are had more directiveness.
Need to prove; One of ordinary skill in the art will appreciate that all or part of flow process that realizes in the foregoing description method; Be to instruct relevant hardware to accomplish through computer program; Described program can be stored in the computer read/write memory medium, and this program can comprise the flow process like the embodiment of above-mentioned each side method when carrying out.Wherein, described storage medium can be magnetic disc, CD, read-only storage memory body (Read-Only Memory, ROM) or at random store memory body (Random Access Memory, RAM) etc.
More than a kind of information-pushing method and device that the application provided have been carried out detailed introduction; Used specific embodiment among this paper the application's principle and embodiment are set forth, the explanation of above embodiment just is used to help to understand the application's method and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to the application's thought, the part that on embodiment and range of application, all can change, in sum, this description should not be construed as the restriction to the application.
Claims (10)
1. an information-pushing method is characterized in that, comprising:
From the problem that the user submits to, extract keyword;
From user's historical behavior data, obtain releasing news of said keyword according to the text matches method;
From user's historical behavior data, add up and the said relevant associated user that releases news according to statistical analysis technique;
Calculate the matching degree between said problem and the said associated user, said problem is pushed to the user of the preset number that begins from the highest associated user of matching degree.
2. method according to claim 1 is characterized in that, the matching degree between said problem of said calculating and the said associated user, and the associated user who said problem is pushed to the preset number that begins from the highest associated user of matching degree comprises:
According to keyword importance keyword assignment weight for extracting to problem in the semanteme statement, wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big;
Matching degree according to releasing news with keyword is that releasing news of obtaining assigns weight, and wherein, the weight that release news lower than matching degree that release news that matching degree is high is big;
The product of the weight of calculating keyword and the weight that releases news of said keyword; With matching degree and the summation of said product value between the classification of problem and the classification that releases news, obtain said problem and first matching degree of issuing between the said user who releases news again;
Order according to from high to low sorts to said first matching degree, and problem is sent to since the highest user user, preset number of first matching degree.
3. method according to claim 1 is characterized in that, the matching degree between said problem of said calculating and the said associated user, and the associated user who said problem is pushed to the preset number that begins from the highest associated user of matching degree comprises:
According to keyword importance keyword assignment weight for extracting to problem in the semanteme statement, wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big;
Matching degree according to releasing news with keyword is that releasing news of obtaining assigns weight, and wherein, the weight that release news lower than matching degree that release news that matching degree is high is big;
From user's historical behavior data, add up the said user who releases news of feedback to the said degree of feedback that releases news according to statistical analysis technique;
Calculate weight, the weight that releases news of said keyword and the product of the said degree of feedback of keyword; With matching degree and the summation of said product value between the classification of problem and the classification that releases news, obtain said problem and second matching degree of feeding back between the said user who releases news again;
Order according to from high to low will sort to said second matching degree, and problem is sent to since the user of the highest user's of second matching degree preset number.
4. method according to claim 1 is characterized in that, the matching degree between said problem of said calculating and the said user, and the associated user who said problem is pushed to the highest associated user's of matching degree preset number comprises:
According to keyword importance keyword assignment weight for extracting to problem in the semanteme statement, wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big;
Matching degree according to releasing news with keyword is that releasing news of obtaining assigns weight, and wherein, the weight that release news lower than matching degree that release news that matching degree is high is big;
From user's historical behavior data, add up the said user who releases news of feedback to the said degree of feedback that releases news according to statistical analysis technique;
The product of the weight of calculating keyword and the weight that releases news of said keyword; Again with matching degree and the summation of said product value between the classification of problem and the classification that releases news; Obtain first matching degree between said problem and the said user who releases news of feedback; And; Calculate weight, the weight that releases news of said keyword and the product of the said degree of feedback of keyword, with matching degree and the summation of said product value between the classification of problem and the classification that releases news, obtain said problem and issue second matching degree between the said user who releases news again;
Order according to from high to low sorts to said first matching degree and second matching degree, and problem is sent to the user that begin, preset number from the highest user of matching degree.
5. according to claim 3 or 4 described methods; It is characterized in that; Saidly from user's historical behavior data, add up the said user who releases news of feedback according to statistical analysis technique and be: the product of the ageing factor that calculates the said user's who releases news of feedback feedback number of times and feedback the said degree of feedback that releases news; Wherein, the ageing factor of the current timing statistics of the near feedback ratio of the current timing statistics of distance distance feedback far away is big.
6. an information push-delivery apparatus is characterized in that, comprising:
Extraction unit is used for extracting keyword from the problem that the user submits to;
Acquiring unit is used for obtaining releasing news of said keyword according to the text matches method from user's historical behavior data;
Statistic unit is used for according to the historical behavior data statistics and said release news relevant associated user of statistical analysis technique from the user;
Push unit is used to calculate the matching degree between said problem and the said associated user, said problem is pushed to the user of the preset number that begins from the highest associated user of matching degree.
7. device according to claim 6 is characterized in that, said push unit comprises:
The first weight allocation subelement is used for according to keyword importance keyword assignment weight for extracting to problem in the semanteme statement, and wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big;
The second weight allocation subelement, being used for according to the matching degree that releases news with keyword is that releasing news of obtaining assigns weight, wherein, the weight that release news lower than matching degree that release news that matching degree is high is big;
The first matching degree computation subunit; Be used to calculate the product of the weight that releases news of weight and the said keyword of keyword; With matching degree and the summation of said product value between the classification of problem and the classification that releases news, obtain said problem and first matching degree of issuing between the said user who releases news again;
First sends subelement, is used for according to from high to low order said first matching degree being sorted, and problem is sent to since the highest user user, preset number of first matching degree.
8. device according to claim 6 is characterized in that, said push unit comprises:
The first weight allocation subelement is used for according to keyword importance keyword assignment weight for extracting to problem in the semanteme statement, and wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big;
The second weight allocation subelement, being used for according to the matching degree that releases news with keyword is that releasing news of obtaining assigns weight, wherein, the weight that release news lower than matching degree that release news that matching degree is high is big;
Degree of feedback statistics subelement is used for feeding back the said user who releases news to the said degree of feedback that releases news according to statistical analysis technique from user's historical behavior data statistics;
The second matching degree computation subunit; Be used to calculate the weight that releases news of the weight of keyword, said keyword and the product of the said degree of feedback; With matching degree and the summation of said product value between the classification of problem and the classification that releases news, obtain said problem and second matching degree of feeding back between the said user who releases news again;
Second sends subelement, is used for will sorting to said second matching degree according to from high to low order, and problem is sent to since the highest user user, preset number of second matching degree.
9. device according to claim 6 is characterized in that, said pushing unit comprises:
The first weight allocation subelement is used for according to keyword importance keyword assignment weight for extracting to problem in the semanteme statement, and wherein, the weight of the keyword that the keyword that importance is high is lower than importance is big;
The second weight allocation subelement, being used for according to the matching degree that releases news with keyword is that releasing news of obtaining assigns weight, wherein, the weight that release news lower than matching degree that release news that matching degree is high is big;
Degree of feedback statistics subelement is used for feeding back the said user who releases news to the said degree of feedback that releases news according to statistical analysis technique from user's historical behavior data statistics;
The COMPREHENSIVE CALCULATING subelement; Be used to calculate the product of the weight that releases news of weight and the said keyword of keyword; Again with matching degree and the summation of said product value between the classification of problem and the classification that releases news; Obtain first matching degree between said problem and the said user who releases news of feedback, and, weight, the weight that releases news of said keyword and the product of the said degree of feedback of calculating keyword; With matching degree and the summation of said product value between the classification of problem and the classification that releases news, obtain said problem and second matching degree of issuing between the said user who releases news again;
The 3rd sends subelement, is used for according to from high to low order said first matching degree and second matching degree being sorted, and problem is sent to the user of the preset number that begins from the highest user of matching degree.
10. according to Claim 8 or 9 described devices, it is characterized in that said degree of feedback statistics subelement comprises:
Computation subunit is used to calculate the said user's who releases news of feedback the feedback number of times and the product of the ageing factor of feedback, and wherein, the ageing factor of the feedback that the near feedback ratio of the current timing statistics of distance is far away apart from current timing statistics is big.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110138313.XA CN102799589B (en) | 2011-05-25 | 2011-05-25 | A kind of information-pushing method and device |
| HK13101053.7A HK1174110A1 (en) | 2011-05-25 | 2013-01-24 | Method and device for information pushing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110138313.XA CN102799589B (en) | 2011-05-25 | 2011-05-25 | A kind of information-pushing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102799589A true CN102799589A (en) | 2012-11-28 |
| CN102799589B CN102799589B (en) | 2016-05-11 |
Family
ID=47198701
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110138313.XA Active CN102799589B (en) | 2011-05-25 | 2011-05-25 | A kind of information-pushing method and device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN102799589B (en) |
| HK (1) | HK1174110A1 (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103870978A (en) * | 2012-12-17 | 2014-06-18 | 北京千橡网景科技发展有限公司 | Intelligent push method and device based on user attribute combination difference |
| CN104699696A (en) * | 2013-12-05 | 2015-06-10 | 深圳市腾讯计算机系统有限公司 | File recommendation method and device |
| CN105099853A (en) * | 2014-04-25 | 2015-11-25 | 国际商业机器公司 | Erroneous message sending preventing method and system |
| CN106294058A (en) * | 2015-05-28 | 2017-01-04 | 腾讯科技（深圳）有限公司 | The target strategy processing problems of operation document determines method and device |
| CN106557516A (en) * | 2015-09-29 | 2017-04-05 | 北京国双科技有限公司 | Data push method and device |
| CN106776542A (en) * | 2016-11-23 | 2017-05-31 | 北京小米移动软件有限公司 | The crucial word treatment method of field feedback, device and server |
| CN107526778A (en) * | 2017-07-22 | 2017-12-29 | 长沙兔子代跑网络科技有限公司 | A kind of method and device that generation race client is excavated according to user behavior data |
| CN110069698A (en) * | 2017-11-01 | 2019-07-30 | 北京京东尚科信息技术有限公司 | Information-pushing method and device |
| CN110377717A (en) * | 2019-07-24 | 2019-10-25 | 江苏满运软件科技有限公司 | Information interacting method, system, equipment and medium based on net connection collaboration |
| CN112182401A (en) * | 2017-11-24 | 2021-01-05 | 创新先进技术有限公司 | Question pushing method and device |
| CN112949305A (en) * | 2021-05-13 | 2021-06-11 | 平安科技（深圳）有限公司 | Negative feedback information acquisition method, device, equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101257512A (en) * | 2008-02-02 | 2008-09-03 | 黄伟才 | Inquiry answer matching method used for inquiry answer system as well as inquiry answer method and system |
| CN101271559A (en) * | 2008-05-16 | 2008-09-24 | 华东师范大学 | Cooperation recommending system based on user partial interest digging |
| CN101520784A (en) * | 2008-02-29 | 2009-09-02 | 富士通株式会社 | Information issuing system and information issuing method |
-
2011
- 2011-05-25 CN CN201110138313.XA patent/CN102799589B/en active Active
-
2013
- 2013-01-24 HK HK13101053.7A patent/HK1174110A1/en unknown
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101257512A (en) * | 2008-02-02 | 2008-09-03 | 黄伟才 | Inquiry answer matching method used for inquiry answer system as well as inquiry answer method and system |
| CN101520784A (en) * | 2008-02-29 | 2009-09-02 | 富士通株式会社 | Information issuing system and information issuing method |
| CN101271559A (en) * | 2008-05-16 | 2008-09-24 | 华东师范大学 | Cooperation recommending system based on user partial interest digging |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103870978A (en) * | 2012-12-17 | 2014-06-18 | 北京千橡网景科技发展有限公司 | Intelligent push method and device based on user attribute combination difference |
| CN104699696A (en) * | 2013-12-05 | 2015-06-10 | 深圳市腾讯计算机系统有限公司 | File recommendation method and device |
| CN104699696B (en) * | 2013-12-05 | 2018-12-28 | 深圳市腾讯计算机系统有限公司 | File recommendation method and device |
| US10044664B2 (en) | 2014-04-25 | 2018-08-07 | International Business Machines Corporation | Prevention of sending messages by mistake |
| CN105099853A (en) * | 2014-04-25 | 2015-11-25 | 国际商业机器公司 | Erroneous message sending preventing method and system |
| CN106294058A (en) * | 2015-05-28 | 2017-01-04 | 腾讯科技（深圳）有限公司 | The target strategy processing problems of operation document determines method and device |
| CN106557516A (en) * | 2015-09-29 | 2017-04-05 | 北京国双科技有限公司 | Data push method and device |
| CN106776542A (en) * | 2016-11-23 | 2017-05-31 | 北京小米移动软件有限公司 | The crucial word treatment method of field feedback, device and server |
| CN106776542B (en) * | 2016-11-23 | 2020-03-03 | 北京小米移动软件有限公司 | Keyword processing method and device for user feedback information and server |
| CN107526778A (en) * | 2017-07-22 | 2017-12-29 | 长沙兔子代跑网络科技有限公司 | A kind of method and device that generation race client is excavated according to user behavior data |
| CN110069698A (en) * | 2017-11-01 | 2019-07-30 | 北京京东尚科信息技术有限公司 | Information-pushing method and device |
| CN110069698B (en) * | 2017-11-01 | 2021-09-03 | 北京京东尚科信息技术有限公司 | Information pushing method and device |
| CN112182401A (en) * | 2017-11-24 | 2021-01-05 | 创新先进技术有限公司 | Question pushing method and device |
| CN112182401B (en) * | 2017-11-24 | 2024-03-29 | 创新先进技术有限公司 | Question pushing method and device |
| CN110377717A (en) * | 2019-07-24 | 2019-10-25 | 江苏满运软件科技有限公司 | Information interacting method, system, equipment and medium based on net connection collaboration |
| CN112949305A (en) * | 2021-05-13 | 2021-06-11 | 平安科技（深圳）有限公司 | Negative feedback information acquisition method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102799589B (en) | 2016-05-11 |
| HK1174110A1 (en) | 2013-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102799589A (en) | Information pushing method and device | |
| CN109033387B (en) | Internet of things searching system and method fusing multi-source data and storage medium | |
| CN110209808B (en) | Event generation method based on text information and related device | |
| CN102110170B (en) | System with information distribution and search functions and information distribution method | |
| CN106960030B (en) | Information pushing method and device based on artificial intelligence | |
| CN102591917A (en) | Data processing method and system and related device | |
| CN101101605A (en) | Method, device and system for searching web page and device for establishing index database | |
| CN102236710A (en) | Method and equipment for displaying news information in query result | |
| CN103870538A (en) | Method, user modeling equipment and system for carrying out personalized recommendation for users | |
| CN103092943A (en) | Method of advertisement dispatch and advertisement dispatch server | |
| CN103514181A (en) | Searching method and device | |
| CN102567423B (en) | Method and system for associated search of poetry | |
| CN103530364A (en) | Method and system for providing download link | |
| CN1741013A (en) | Automatic adaptive system for customer service and its method | |
| CN111191111A (en) | Content recommendation method, device and storage medium | |
| CN103389976A (en) | Searching method and searching system for terminal | |
| CN103064842A (en) | Information subscription processing device and information subscription processing method | |
| CN104462347B (en) | The sorting technique and device of keyword | |
| CN103927339A (en) | System and method for reorganizing knowledge | |
| CN112825089A (en) | Article recommendation method, article recommendation device, article recommendation equipment and storage medium | |
| CN109145092B (en) | Database updating and intelligent question and answer management method, device and equipment | |
| CN107168966A (en) | A kind of search engine index construction method and device | |
| CN105512270B (en) | Method and device for determining related objects | |
| CN102760127A (en) | Method, device and equipment for determining resource type based on extended text information | |
| KR20220061388A (en) | A recording medium in which the program providing the keyword-item mapping information service of news articles |
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: 1174110 Country of ref document: HK |
|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: GR Ref document number: 1174110 Country of ref document: HK |