CN103365899A - Question recommending method and question recommending system both in questions-and-answers community - Google Patents
Question recommending method and question recommending system both in questions-and-answers community Download PDFInfo
- Publication number
- CN103365899A CN103365899A CN2012100959175A CN201210095917A CN103365899A CN 103365899 A CN103365899 A CN 103365899A CN 2012100959175 A CN2012100959175 A CN 2012100959175A CN 201210095917 A CN201210095917 A CN 201210095917A CN 103365899 A CN103365899 A CN 103365899A
- Authority
- CN
- China
- Prior art keywords
- user
- weight
- key
- field
- issue word
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention is applicable to the technical field of internet technology, provides a question recommending method and a question recommending system both in a questions-and-answers community. The method includes the following steps: matching a question keyword list of a to-be-recommended question with a preset weighted question keyword list of each user to acquire a weight sum of successfully-matched question keywords corresponding to each user, matching the question keyword list with preset weighted field keyword lists of each user to acquire a maximum value of a weight sum of successfully-matched field question keywords corresponding to each user, combining the weight sum of the successfully-matched question keywords corresponding to each user and the maximum value of the weight sum of the field question keywords to acquire relevancy of the to-be-recommended question and a user, and pushing the to-be-recommended question to one or more users with highest relevancy to improve answering rate of the users to the questions in the questions-and-answers community and improve user experience effect.
Description
Technical field
The invention belongs to Internet technical field, relate in particular to method for recommending problem and system in a kind of Ask-Answer Community.
Background technology
At present, in the internet Ask-Answer Community, often need to automatically recommend problem to the user, existing recommend method mainly is divided into two classes, i.e. collaborative recommendation method and based on the recommend method of attribute.In collaborative recommendation method, the user that system searching is similar to user to be recommended, and according to these users selected project once, give this user with suitable project recommendation; Perhaps with reference to the project similar to project to be recommended, and according to the user who once selected these projects, for suitable user recommends this project.Can find out him according to user's historical buying behavior such as electronic emporium may interested commodity and be that he recommends.And in the method based on attribute, can recommend according to some attribute of user, such as, can be user's recommended dietary according to user's height and body weight, perhaps recommend finance product according to user's financial position for it.
Yet, in Ask-Answer Community, a user proposes own interested problem, and answered for it by the user that can answer this problem, because problem is no longer needed to understand the user of this problem and is continued answer after answering, therefore seldom have problem of the common answer of a plurality of users, seldom exist in other words a problem to be answered by a plurality of users, therefore the method based on Collaborative Recommendation is not suitable for this situation.And more be applicable to question recommending based on the method that user property is recommended.In Ask-Answer Community, user's attribute generally comprises the problem quantity that he answered, his active degree, and the length of problem etc.But, present method for recommending problem or lay particular emphasis on to the user and push the problem that he answered can be answered this problem to guarantee him; He perhaps lays particular emphasis on the problem of recommending him not answer to the user as far as possible, so that can not be fed up with.The former can cause same problem always to be received by same user, so that system can't obtain other people answer, and pushed user is sick of, and latter causes the user to receive too much his unanswerable problem.
Summary of the invention
The purpose of the embodiment of the invention is to provide method for recommending problem and the system in a kind of Ask-Answer Community, be intended to solve in Ask-Answer Community when the user carried out question recommending, single and the situation that pushes the uninterested problem of user of the problem that existence is recommended to the user causes in the Ask-Answer Community the not high and not good problem of user's experience effect of response rate to problem.
The embodiment of the invention is achieved in that the method for recommending problem in a kind of Ask-Answer Community, and described method comprises the steps:
Obtain the key to the issue word tabulation of problem to be recommended;
Described key to the issue word tabulation and each user's who presets Weight key to the issue word tabulation are mated, obtain the weight sum with the corresponding key to the issue word that the match is successful of described each user;
The tabulation of described key to the issue word is mated with all Weight field lists of keywords of default each user respectively, obtain the maximal value in the weight sum of all field keywords that the match is successful corresponding with described each user;
In conjunction with the maximal value in the weight sum of the weight sum of the key to the issue word that the match is successful corresponding to same user and field keyword, obtain described problem to be recommended and described user's the degree of correlation;
According to described problem to be recommended and each user's the degree of correlation, described problem to be recommended is pushed to one or more the highest user of the degree of correlation.
Another purpose of the embodiment of the invention is to provide the system of the question recommending in a kind of Ask-Answer Community, and described system comprises:
Lists of keywords is obtained master unit, is used for obtaining the key to the issue word tabulation of problem to be recommended;
Weight sum acquiring unit is used for described key to the issue word tabulation and each user's who presets Weight key to the issue word tabulation are mated, and obtains the weight sum with the corresponding key to the issue word that the match is successful of described each user;
The maximal value acquiring unit, be used for described key to the issue word tabulation is mated with all Weight field lists of keywords of default each user respectively, obtain the maximal value in the weight sum of all field keywords that the match is successful corresponding with described each user;
Degree of correlation acquiring unit is used for the maximal value in conjunction with the weight sum of the weight sum of the key to the issue word that the match is successful corresponding to same user and field keyword, obtains described problem to be recommended and described user's the degree of correlation; And
The problem push unit is used for the degree of correlation according to described problem to be recommended and each user, and described problem to be recommended is pushed to one or more the highest user of the degree of correlation.
The embodiment of the invention is by mating the key to the issue word tabulation of problem to be recommended with each user's who presets Weight key to the issue word tabulation, obtain the weight sum of the key to the issue word that the match is successful corresponding to this each user, this key to the issue word tabulation is mated with all Weight field lists of keywords of each user who presets respectively, obtain the maximal value in the weight sum of all field keywords that the match is successful corresponding to this each user, and the maximal value in the weight sum of the weight sum of the combination key to the issue word that the match is successful corresponding to same user and field keyword, obtain this problem to be recommended and this user's the degree of correlation, this problem to be recommended is pushed to one or more the highest user of the degree of correlation, solved in Ask-Answer Community when the user carried out question recommending, single and the situation that pushes the uninterested content of user of the problem that existence is recommended to the user, cause in the Ask-Answer Community the not high and not good problem of user's experience effect of response rate to problem, avoided the problem of recommending to the user single and belong to the uninterested problem of user, thereby improved the response rate of user to problem in the Ask-Answer Community, also promoted the experience effect of user in Ask-Answer Community.
Description of drawings
Fig. 1 is the realization flow figure of the method for recommending problem in the Ask-Answer Community that provides of first embodiment of the invention;
Fig. 2 is the realization flow figure of the Weight key to the issue word tabulating method that obtains the user that provides of second embodiment of the invention;
Fig. 3 is the structural drawing of the question recommending system in the Ask-Answer Community that provides of third embodiment of the invention;
Fig. 4 is the concrete structure figure of the maximal value acquiring unit that provides of third embodiment of the invention.
Embodiment
In order to make purpose of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, is not intended to limit the present invention.
The embodiment of the invention reaches all Weight field lists of keywords of each default user by the Weight key to the issue word tabulation that utilizes each default user, obtain the maximal value in the weight sum of all field keywords that the match is successful corresponding to the weight sum of the key to the issue word that the match is successful corresponding to this each user and this each user, in conjunction with the maximal value in the weight sum of the weight sum of the key to the issue word that the match is successful corresponding to same user and field keyword, obtain this problem to be recommended and this user's the degree of correlation, this degree of correlation is sorted from high to low, user to the high default number of the degree of correlation carries out the recommendation of problem, thereby when in Ask-Answer Community, the user being carried out question recommending, can recommend more exactly the interested problem of user etc., so that user's experience is better.
Below in conjunction with specific embodiment specific implementation of the present invention is described in detail:
Embodiment one:
Fig. 1 shows the realization flow of the method for recommending problem in the Ask-Answer Community that first embodiment of the invention provides, and details are as follows:
In step S101, obtain the key to the issue word tabulation of problem to be recommended.
In specific implementation process, after having new problem to propose, wish that the user that can answer this problem answers this new problem, yet owing to not knowing which field and/or which user etc. can answer this new problem, perhaps can't learn that its problem that can answer is suggested owing to can answer the user of this new problem, the method for recommending problem that then utilizes the embodiment of the invention to provide can obtain first the lists of keywords of the problem of this proposition, be in particular in the problem of extracting this proposition, can reflect the keyword of this problem content information etc., finally this new problem can be recommended suitable user's answer according to this lists of keywords.Wherein, this key to the issue word tabulation is the set of the keyword relevant with the content of this problem to be recommended, also namely should the relevant content information of keyword from reflecting that this problem to be recommended comprises.
In step S102, this key to the issue word tabulation and each user's who presets Weight key to the issue word tabulation are mated, obtain the weight sum of the key to the issue word that the match is successful corresponding to this each user.
Before execution in step S102, the method for recommending problem in this Ask-Answer Community also comprises:
Obtain the problem list that the user once answered and refused to answer, with this problem list according to time sequence, extract occurrence number in the problem list after the ordering greater than 1 key to the issue word, and generate this user's key to the issue word tabulation;
According to the Weight Acquisition relation of default key to the issue word, obtain this user's Weight key to the issue word tabulation.
Wherein, specific implementation process such as the following embodiment two of the step that the above-mentioned Weight key to the issue word that obtains each user is tabulated do not repeat them here.
This step S102 specifically comprises:
The Weight key to the issue word tabulation of a user among the tabulation of the key to the issue word of this problem to be recommended and default all users is mated, obtain and be present in simultaneously the key to the issue word that the match is successful in this lists of keywords and the tabulation of this Weight key to the issue word;
Based on this Weight key to the issue word tabulation, obtain weight corresponding to key to the issue word that the match is successful with this;
The weight sum with the corresponding key to the issue word that the match is successful of this user is obtained in the weight addition that this is corresponding with this key to the issue word that the match is successful;
Upgrade this user for the next user among all default users, repeat above-mentioned steps, until obtain with these all default users in the weight sum of the corresponding key to the issue word that the match is successful of each user after, withdraw from.
As one embodiment of the invention, the key to the issue word tabulation of supposing problem to be recommended is { a1, a2, a3}, these all default users are 3 users, are respectively U1, U2, U3, and the Weight key to the issue word tabulation that U1 is corresponding is { U11, T11, U12, T12 ... }, the Weight key to the issue word tabulation that U2 is corresponding is { U21, T21, U22, T22, ..., the Weight key to the issue word tabulation that U3 is corresponding is { U31, T31, U32, T32, ..., wherein U11 etc. is keyword, and T11 is the weight of keyword U11, through after the above-mentioned coupling:
For U1, the key to the issue word tabulation of this problem to be recommended is mated with the Weight key to the issue word tabulation of U1, the U11 that supposes to obtain in the Weight key to the issue word tabulation of a1 and U1 in the key to the issue word tabulation of this problem to be recommended mates, and the weight sum of corresponding with the U1 so key to the issue word that the match is successful is T11;
For U2, the key to the issue word tabulation of this problem to be recommended is mated with the Weight key to the issue word tabulation of U2, suppose to obtain a1, a2 U23, the U24 coupling in the Weight key to the issue word tabulation of corresponding and U2 respectively in the key to the issue word tabulation of this problem to be recommended, the weight sum of corresponding with the U2 so key to the issue word that the match is successful is (T23+T24);
For U3, the key to the issue word tabulation of this problem to be recommended is mated with the Weight key to the issue word tabulation of U3, suppose to obtain the a1 in the key to the issue word tabulation of this problem to be recommended, a2, U33, U35 among a3 difference correspondence and the U3, U36 coupling, the weight sum of corresponding with the U2 so key to the issue word that the match is successful is (T33+T35+T36).
In embodiments of the present invention, the tabulation of this user's Weight key to the issue word has reflected that the user answers or refusal answer have this keyword the frequency of problem, thereby the method for recommending problem among use the present invention calculates the weight sum of the key to the issue word that the match is successful, weighed the wish that the user answers this problem.When the user repeatedly answers when being with a certain crucial word problem, this keyword can have a high weight in this user's lists of keywords; Otherwise, when the user repeatedly refuses to answer when being with a certain crucial word problem, the weight of this word will reduce until-1, in the matching degree of the key to the issue word tabulation of obtaining problem to be recommended and each user or after claiming weight, can be hopeful in other words greatly the user that answers to question recommending to the weight sum.
In step S103, the tabulation of this key to the issue word is mated with all Weight field lists of keywords of default each user respectively, obtain the maximal value in the weight sum of all field keywords that the match is successful corresponding with this each user.
Before execution in step S103, the method for recommending problem in this Ask-Answer Community also comprises:
Obtain in all problems of designated field occurrence number greater than the set of the field keyword of default value;
Obtain the first frequency that each keyword in this set occurs in all problems of this designated field, and obtain the second frequency that this each keyword occurs in all problems of all spectra;
According to this each keyword first frequency and this second frequency, obtain the weight of each keyword of this designated field, and generate the Weight field lists of keywords of this designated field.
In specific implementation process, if for designated field c, be extracted among the c occurrence number more than or equal to the set W={w1 of the field keyword of a certain numerical value, w2, ... wn}, this numerical value can be chosen according to actual conditions, such as can getting 10 integers such as grade, this field keyword refer to this designated field in the set of the keyword that the content of all problems is relevant among the c, also namely should relevant keyword content information that all problems comprises from reflect this field c, to each keyword wi among the W, i=1,2 ..., n, carry out following calculating:
The Issue Totals that note was delivered at all spectra is N, this all spectra comprises field a, b, c etc., remember that the problem number that comprises keyword wi in N the problem is di, remember that the Issue Totals among a certain field c is Nc, remember that the problem number that comprises keyword wi among the c of this field is Nci, then distribute a weight parameter cwi for keyword wi, obtain cwi according to following formula (1), wherein, Nci/Nc represents the first frequency that each keyword occurs in all problems of this designated field, di/N represents the second frequency that this each keyword occurs in all problems of all spectra
cwi＝log((Nci/Nc)×(N/di))×log(N/di)，(1)
If wi is then removed in cwi≤0 from W, otherwise, this wi kept.
Afterwards, can also carry out normalized to the weight wi of each keyword, obtain the Weight field lists of keywords of this designated field c after the renewal, also be after each keyword among the pair set W as above calculates, get the maximal value of the cwi of all words among the set W, be designated as max_cw, and all cwi are carried out the normalization computing: represent normalization cwi before, the then cwi=old-_cwi/max_cw after the normalization with old_cwi.All word wi among the set W and associated weight cwi consist of the Weight lists of keywords of field c.
Comprehensive above-mentioned steps then can be obtained the Weight lists of keywords in each field in all spectra, also can be called Weight field lists of keywords, and this Weight field lists of keywords and sole user are irrelevant, but relevant with all users.
Wherein, the weight of each keyword in this Weight field lists of keywords be scope (0,1] in real number, represent the therewith relevant degree in field of certain keyword.
This step S103 specifically comprises:
A. obtain the related field of all problems that a user among all default users answered;
B. according to the Weight field lists of keywords in each field in the default all spectra, obtain all Weight field lists of keywords of this user corresponding with this related field;
C. key to the issue word tabulation that will this problem to be recommended is mated with a Weight field lists of keywords in all Weight field lists of keywords of this user, obtains to be present in simultaneously the field keyword that the match is successful in this key to the issue word tabulation and this Weight field lists of keywords;
D. based on this Weight field lists of keywords, obtain the weight sum with the corresponding field keyword that the match is successful of this Weight field lists of keywords;
E. a Weight field lists of keywords of upgrading in all Weight field lists of keywords of this user is the next Weight field lists of keywords in all Weight field lists of keywords of this user, repeat above-mentioned steps c to d, until obtain with this user the weight sum of the corresponding field keyword that the match is successful of each Weight field lists of keywords after, execution in step f;
F. obtain the maximal value in this weight sum;
G. upgrade this user for the next user among all default users, repeat above-mentioned steps a to f, until obtain with these all default users in the weight sum of the corresponding field keyword that the match is successful of each user in maximal value.
As one embodiment of the invention, suppose that the key to the issue word tabulation of problem to be recommended is { a1, a2, a3}, default all spectra are { q1, q2, q3, q4}, the field lists of keywords that q1 is corresponding is { U11, T11, U12, T12 ... }, the field lists of keywords that q2 is corresponding is { U21, T21, U22, T22, ..., the field lists of keywords that q3 is corresponding is { U31, T31, U32, T32, ..., the field lists of keywords that q4 is corresponding is { U41, T41, U42, T42 ... }, should all default users be 3 users, be respectively U1, U2, U3, and the related field of the hypothesis U1 all problems of answering is { q2}, and the related field of all problems that U2 answered is { q1, q3}, and the related field of all problems that U3 answered is { q2, q3, q4}, wherein U11 etc. is keyword, T11 etc. are the weight of keyword U11, then pass through above-mentioned coupling after
For U1, the related field of all problems that U1 was answered is that { q2} is { q1 with all spectra of presetting, q2, q3, q4} mates, then corresponding Weight field lists of keywords is the corresponding Weight of q2 field lists of keywords { U21, T21, U22, T22, ..., the U11 that supposes to obtain in the Weight key to the issue word tabulation of a1 and U1 in the key to the issue word tabulation of this problem to be recommended mates, the weight sum of corresponding with the U1 so key to the issue word that the match is successful is T11, and then the maximal value of the weight sum of the corresponding field keyword that the match is successful of Weight field lists of keywords of U1 also is T11;
For U2, the related field of all problems that U2 was answered is { q1, q3} is { q1 with all spectra of presetting, q2, q3, q4} mates, then corresponding coupling field is respectively q1, q3, q1, the corresponding Weight of q3 field lists of keywords is { U11 respectively, T11, U12, T12, ..., { U31, T31, U32, T32, ..., suppose to obtain the a1 in the key to the issue word tabulation of this problem to be recommended, the Weight key to the issue word tabulation { U11 of a2 difference correspondence and U2, T11, U12, T12, ... in U13, the U14 coupling, and the a1 difference in the tabulation of the key to the issue word of this problem to be recommended is corresponding and the Weight key to the issue word tabulation { U31 of U2, T31, U32, T32 ... } in U33, the U34 coupling, the weight sum of corresponding with the U2 so key to the issue word that the match is successful has two, be respectively (T13+T14), (T33+T34), if further judgement is learnt (T13+T14) greater than (T33+T34), then the maximal value of the weight sum of the corresponding field keyword that the match is successful of Weight field lists of keywords of U2 also is (T13+T14);
For U3, the related field of all problems that U3 was answered is { q2, q3, q4} is { q1 with all spectra of presetting, q2, q3, q4} mates, then corresponding coupling field is respectively q2, q3, q4, q2, q3, the corresponding Weight of q4 field lists of keywords is { U21 respectively, T21, U22, T22, ..., { U31, T31, U32, T32, ..., { U41, T41, U42, T42, ..., the key to the issue word tabulation of this problem to be recommended is mated with the Weight key to the issue word tabulation of U3, suppose to obtain the a2 Weight key to the issue word tabulation { U11 of corresponding and U3 respectively in the key to the issue word tabulation of this problem to be recommended, T11, U12, T12 ... } in U13 coupling, and the a3 difference in the tabulation of the key to the issue word of this problem to be recommended is corresponding and the Weight key to the issue word tabulation { U21 of U3, T21, U22, T22 ... } in U23, U24 coupling, and the a1 in the tabulation of the key to the issue word of this problem to be recommended, the Weight key to the issue word tabulation { U31 of a3 difference correspondence and U3, T31, U32, T32 ... } in U33, U35, the U36 coupling, the weight sum of corresponding with the U2 so key to the issue word that the match is successful is (T13), (T23+T24), (T33+T35+T36), if further judge learn (T33+T35+T36) greater than (T23+T24) greater than (T13), then the maximal value of the weight sum of the corresponding field keyword that the match is successful of Weight field lists of keywords of U3 also is (T33+T35+T36).
Thereby, through above-mentioned steps, obtained with these all default users in the weight sum of the corresponding field keyword that the match is successful of each user in maximal value.
In embodiments of the present invention, this Weight field lists of keywords has reflected the temperature situation of keyword in a certain field, after execution in step S104, for a certain user, if it is maximum to have obtained under the related problem of this user in the field weight sum in a corresponding field, illustrated that then this user is even without the problem of answering this field, also probably can answer the problem in this field, like this with regard to so that may recommend the people who answers in history outside the maximum people of this problem with the recommendation problem, also i.e. this user, also so that a problem has more candidate to recommend the user, the user has more selection problem.
In step S104, in conjunction with the maximal value in the weight sum of the weight sum of the key to the issue word that the match is successful corresponding to same user and field keyword, obtain this problem to be recommended and this user's the degree of correlation.
Wherein, this step S104 specifically comprises:
Maximal value in the weight sum of the weight sum of the corresponding key to the issue word that the match is successful of same user and field keyword is carried out addition;
With the numerical value of gained after the addition degree of correlation as this problem to be recommended and this user.
In step S105, according to this problem to be recommended and each user's the degree of correlation, this problem to be recommended is pushed to one or more the highest user of the degree of correlation.
In specific implementation process, obtain after this problem to be recommended and each user's the degree of correlation, if there are a plurality of users, then can obtain a plurality of degrees of correlation corresponding with these a plurality of users, can sort from high to low to these a plurality of degrees of correlation, obtain several users corresponding to the degree of correlation that rank in the top, give these several users with the question recommending of recommendation to be with, user's number that this Receiver Problem is recommended can be set according to actual needs.
In embodiments of the present invention, when based on this problem to be recommended, the user's who obtains key to the issue word tabulation score or title weight sum are lower, and user's field lists of keywords score or when claiming the maximal value of weight sum higher, illustrate that this user answered and problem like these question marks to be recommended, and field and the affiliated domain-specific of this problem to be recommended under this similar problem, represent this user there is a strong possibility to answer this problem to be recommended, like this with regard to so that problem may be recommended the people who answers in history outside the maximum people of this problem, so that a problem has more candidate to recommend the user, also make the user that more selection problem is arranged.
In embodiments of the present invention, the problem that each is to be recommended and user's the degree of correlation are combined into by field lists of keywords and this matching degree with the recommendation problem in the related field of the tabulation of this user's Weight key to the issue word and this problem of answering with the matching degree of recommendation problem and this user, so that when recommending problem to the user, avoided same user to receive same problem, and can't obtain other users' answer answer, also avoided the user to receive too much his unanswerable problem and be fed up with etc., such as, owing to having introduced the negative weight of user's keyword, the method can when pushing fresh problem to the user, avoid pushing his uninterested problem.If the user does not answer certain class problem, keyword weight wherein will reduce gradually, thereby system will can for the user recommends such problem, because the method for this recommendation problem uses the field antistop list, also can not attempt recommending for the user other problem in this field.
Embodiment two:
The realization flow of the method that the Weight key to the issue word that obtains the user that Fig. 2 shows second embodiment of the invention to be provided is tabulated, before method for recommending problem in utilizing this Ask-Answer Community is recommended new problem to a certain user, need to obtain this user's Weight key to the issue word tabulation, details are as follows:
In step S201, obtain all problems that the user once answered and refused to answer, this all problems is successively sorted by time of origin.
In step S202, extract occurrence number in the problem after the ordering greater than 1 key to the issue word, and form this user's key to the issue word tabulation.
In step S203, the initial weight that each keyword in this key to the issue word tabulation is set is 0.
In specific implementation process, at first collect the problem list Q that a certain user U once answered and refused to answer, the problem among the Q is successively sorted by time of origin, extract occurrence number among the Q greater than the set of 1 key to the issue word, as the tabulation of key to the issue word, this key to the issue word tabulation is set is W={w1, w2 ..., wn}, and be each keyword wi, i=1 wherein, 2, ..., n arranges a real parameters or claims initial weight Ulrwi, i=1 wherein, 2 ..., n, the Ulrwi initial value is 0, and a learning rate constant RATE is set, can learns that according to practical experience the span of RATE is (0,1), such as, RATE=0.02 can be set.
In step S204, obtain the problem label of first problem in all problems after this ordering and all crucial word problem parameters of this key to the issue word tabulation.
In step S205, in conjunction with the initial weight of the problem label of this first problem, these all keywords, problem parameter and according to default Weight Acquisition relation, obtain the first weight of the first keyword in this key to the issue word tabulation.
In step S206, upgrade the initial weight of this first keyword and be the first weight of this first keyword.
In specific implementation process, to the first problem among the Q, carry out following calculating:
Write down the label label of this first problem, if user U has answered this problem, problem label label=1 then, otherwise label=0.If keyword wi appears in this first problem, problem parameter x i=1 then, otherwise xi=0.Then each the keyword wi among the pair set W can obtain corresponding problem parameter, can according to the following equation or claim that default Weight Acquisition relation (2), (3), (4) calculate delta_wi:
delta_wi＝(label-p)×RATE×xi，(4)
For keyword wi, according to assignment method old_Ulrwi=Ulrwi, it is the first weight Ulrwi that Ulrwi=old_Ulrwi+delta_wi upgrades with the corresponding initial weight of keyword wi.
In step S207, judged whether to travel through all keywords in this key to the issue word tabulation, be, then go to execution in step S209, no, execution in step S208 then.
In step S208, with the next keyword of this key to the issue word tabulation as the first keyword.
In specific implementation process, when first keyword that has upgraded this key to the issue word tabulation in this first problem after the shared weight, need to continue to upgrade the weight of other keywords in this first problem, until upgrade the weight of all keywords in this key to the issue word tabulation, afterwards, continue to upgrade all keywords in this lists of keywords shared weight in next problem.
In step S209, judge whether to have traveled through all problems, be execution in step S211 then, withdraw from afterwards, otherwise execution in step S210.
In step S210, with this ordering after all problems in next problem as first problem.
In step S211, obtain the weight of all keywords in the key to the issue word tabulation after the renewal.
Particularly, utilize above-mentioned steps can obtain the weight of each keyword wi in this first problem among the W.After having upgraded parameter Ulrwi corresponding to each keyword wi, with the initial parameter of the Ulrwi after upgrading as keyword wi in the next problem, also be, to the next problem among the Q, if should this wi occur in the next one problem, then re-execute above-mentioned steps with the Ulrwi value after renewal corresponding to this wi as initial value, continue to upgrade this Ulrwi, until this keyword wi is when certain problem occurs for the last time in Q, the Ulrwi value after the renewal is as the final real parameters value of keyword wi.If certainly to the next problem among the Q, wi no longer occurs, the final real parameters value Ulrwi that then wi is corresponding then is value after last the renewal.Thereby, utilize above-mentioned steps finally can obtain the Weight key to the issue word tabulation ULRW={Ulrw1 of the key to the issue word W of this user U, Ulrw2 ..., Ulrwn}, Ulrwi represent the weight of user U and keyword wi, i=1, and 2 ..., n.
In step S212, according to comprise this tabulation in the frequency that crucial word problem number occurs in all problems, and the shared byte number of this one of them keyword continues to upgrade the weight of all keywords in this tabulation.
In step S213, to exporting after this key to the issue word tabulation normalization after upgrading.
In embodiments of the present invention, also then to each the keyword wi among the W, continue to do following calculating:
Remember that the Issue Totals that whole users delivered is N, the problem number that comprises keyword wi in the problem that all user delivered is di, then distribute a parameter idfi=ln (N/di) for keyword wi, wherein ln is the natural logarithm function, also distribute a parameter b ytei for keyword wi, be designated as the byte number that this keyword takies in calculator memory, this bytei is known, and the associated weight that continues calculating user U and keyword wi is Uwi, Uwi=Ulrwi*idfi*ln (bytei), then can obtain in all problems according to above-mentioned calculating, the associated weight Uwi of user U and each keyword wi is designated as tabulation UW={Uw1, Uw2, ..., Uwn} gets all Uwi among the set UW, i=1,2 ..., the maximal value of the absolute value of n, be designated as max_Uw, and all Uwi are carried out the normalization computing, old_Uwi represents the Uwi before the normalization, then Uuwi=old_Uwi/the max_Uw after the normalization.
To sum up should, the Weight key to the issue word tabulation that all keyword wi among the set W and associated weight Uwi corresponding to this wi have consisted of user U, wherein, the weight of key to the issue word is that scope is [1,1] real number in, accept the degree of this keyword on the occasion of the expression user, negative value represents that then the user refuses the degree of this keyword.
In embodiments of the present invention, according to the user answer and the problem condition of refusal as can be known, if the weighted value that a certain key to the issue word is corresponding is higher, illustrate that then this user has the ability or have a mind to answer to comprise this key to the issue word problem etc., otherwise, this user may be to this problem not too interested even refusal answer and comprise this key to the issue word problem etc., thereby the tabulation of user's Weight key to the issue word has reflected that to a certain extent the user processes the situation of problem in community's question and answer, then managerial personnel of this Ask-Answer Community etc. then can be according to this user's question and answer situation, recommend problem to the user targetedly, so that the user can obtain interested problem in time, also help experience that promotes the user etc.
One of ordinary skill in the art will appreciate that all or part of step that realizes in above-described embodiment method is to come the relevant hardware of instruction to finish by program, the program of being somebody's turn to do can be stored in the computer read/write memory medium, the storage medium that is somebody's turn to do is such as ROM/RAM, disk, CD etc.
Embodiment three:
Fig. 3 shows the structure of the question recommending system in the Ask-Answer Community that third embodiment of the invention provides, and for convenience of explanation, only shows the part relevant with the embodiment of the invention.
The system of the question recommending in this Ask-Answer Community comprises that lists of keywords obtains master unit 31, weight sum acquiring unit 32, maximal value acquiring unit 33, degree of correlation acquiring unit 34 and problem push unit 35, wherein:
Lists of keywords is obtained master unit 31, is used for obtaining the key to the issue word tabulation of problem to be recommended.
In embodiments of the present invention, after having new problem to propose, wish that the user that can answer this problem answers this new problem, yet owing to not knowing which field and/or which user etc. can answer this new problem, perhaps can't learn that its problem that can answer is suggested owing to can answer the user of this new problem, the method for recommending problem that then utilizes the embodiment of the invention to provide can utilize lists of keywords to obtain the lists of keywords that master unit 31 obtains first the problem of this proposition, finally this new problem can be recommended suitable user's answer according to this lists of keywords.Wherein, this key to the issue word tabulation is the set of the keyword relevant with the content of this problem to be recommended, also namely should the relevant content information of keyword from reflecting that this problem to be recommended comprises.
Weight sum acquiring unit 32 is used for this key to the issue word tabulation and each user's who presets Weight key to the issue word tabulation are mated, and obtains the weight sum of the key to the issue word that the match is successful corresponding to this each user.
In embodiments of the present invention, before triggering weight sum acquiring unit 32, the question recommending system in this Ask-Answer Community also comprises:
The tabulation generation unit is used for obtaining the problem list that each user once answered and refused to answer, and with this problem list according to time sequence, extracts occurrence number in the problem list after the ordering greater than 1 key to the issue word, and generates this user's key to the issue word tabulation; And
Weight tabulation acquiring unit is used for the Weight Acquisition relation according to default key to the issue word, obtains this user's Weight key to the issue word tabulation.
Wherein, the weight of key to the issue word is the real number of scope in [1,1], accepts the degree of this keyword on the occasion of the expression user, and negative value represents that then the user refuses the degree of this keyword.And specific implementation process such as the above-mentioned embodiment two of the step that the Weight key to the issue word that obtains each user is tabulated do not repeat them here.
In addition, this weight sum acquiring unit 32 specifically comprises:
Matching keywords acquiring unit 321, be used for will this problem to be recommended the tabulation of key to the issue word mate with default all users' a user's Weight key to the issue word tabulation, obtain and be present in simultaneously this lists of keywords and this Weight key to the issue word key to the issue word that the match is successful in tabulating;
The weight sum is obtained subelement 323, is used for being somebody's turn to do weight addition corresponding to key to the issue word that the match is successful with this, obtains the weight sum with the corresponding key to the issue word that the match is successful of this user; And
Updating block 324, be used for upgrading this user and be next user of all the default users, trigger this matching keywords acquiring unit, until obtain with these all default users in the weight sum of the corresponding key to the issue word that the match is successful of each user after, withdraw from.
In embodiments of the present invention, the tabulation of this user's Weight key to the issue word has reflected that the user answers or refusal answer have this keyword the frequency of problem, thereby the method for recommending problem among use the present invention calculates the weight sum of the key to the issue word that the match is successful, weighed the wish that the user answers this problem.When the user repeatedly answers when being with a certain crucial word problem, this keyword can have a high weight in this user's lists of keywords; Otherwise, when the user repeatedly refuses to answer when being with a certain crucial word problem, the weight of this word will reduce until-1, in the matching degree of the key to the issue word tabulation of obtaining problem to be recommended and each user or after claiming weight, can be hopeful in other words greatly the user who answers to question recommending weight sum.
Maximal value acquiring unit 33, be used for this key to the issue word tabulation is mated with all Weight field lists of keywords of default each user respectively, obtain the maximal value in the weight sum of all field keywords that the match is successful corresponding with this each user.
In embodiments of the present invention, before triggering maximal value acquiring unit 33, the question recommending system in this Ask-Answer Community comprises:
The set acquiring unit is used for obtaining all problems occurrence number of designated field greater than the set of the field keyword of default value;
The frequency acquisition unit is used for obtaining the first frequency that each keyword of this set occurs, and obtains the second frequency that this each keyword occurs in all problems of all spectra in all problems of this designated field; And
Tabulation generation unit in Weight field is used for according to this each keyword first frequency and this second frequency, obtains the weight of each keyword of this designated field, and generates the Weight field lists of keywords of this designated field.
In embodiments of the present invention, utilize set acquiring unit, frequency acquisition unit and Weight field tabulation generation unit can obtain the Weight lists of keywords in each field in all spectra, also can be called Weight field lists of keywords, this Weight field lists of keywords and sole user are irrelevant, but relevant with all users.The weight of each keyword in this Weight field lists of keywords be scope (0,1] in real number, indicate the therewith relevant degree in field of certain keyword.
As shown in Figure 4, this maximal value acquiring unit 33 comprises that specifically relating to field acquiring unit 41, field keyword acquiring unit 42, matching keywords acquiring unit 43, weight sum acquiring unit 44, the first updating block 45, maximal value obtains subelement 46 and the second updating block 47, wherein:
Relate to field acquiring unit 41, be used for obtaining the related field of all problems that a user of all default users answered;
Field keyword acquiring unit 42 is used for the Weight field lists of keywords according to default each field of all spectra, obtains all Weight field lists of keywords of this user corresponding with this related field;
Matching keywords acquiring unit 43, be used for will this problem to be recommended the tabulation of key to the issue word mate with a Weight field lists of keywords of all Weight field lists of keywords of this user respectively, obtain respectively be present in simultaneously that this key to the issue word is tabulated and this Weight field lists of keywords in the field keyword that the match is successful;
Weight sum acquiring unit 44 is used for based on this Weight field lists of keywords, obtains the weight sum with the corresponding field keyword that the match is successful of this Weight field lists of keywords;
The first updating block 45, next Weight field lists of keywords in all Weight field lists of keywords that a Weight field lists of keywords that is used for all Weight field lists of keywords of this user of renewal is this user, trigger this matching keywords acquiring unit 43, until obtain with this user the weight sum of the corresponding field keyword that the match is successful of each Weight field lists of keywords after, trigger maximal value and obtain subelement 46;
Maximal value is obtained subelement 46, is used for obtaining the maximal value of this weight sum; And
The second updating block 47, be used for upgrading this user and be next user of all the default users, trigger this and relate to field acquiring unit 41, until obtain with these all default users in the weight sum of the corresponding field keyword that the match is successful of each user in maximal value.
In embodiments of the present invention, this Weight field lists of keywords has reflected the temperature situation of keyword in a certain field, for a certain user, if it is maximum to have obtained under the related problem of this user in the field weight sum in a corresponding field, illustrated that then this user is even without the problem of answering this field, also probably can answer the problem in this field, like this with regard to so that may recommend the people who answers in history outside the maximum people of this problem with the recommendation problem, also i.e. this user, also so that a problem has more candidate to recommend the user, the user has more selection problem.
Degree of correlation acquiring unit 34 is used for the maximal value in conjunction with the weight sum of the weight sum of the key to the issue word that the match is successful corresponding to same user and field keyword, obtains this problem to be recommended and this user's the degree of correlation.
In embodiments of the present invention, degree of correlation acquiring unit 34 carries out addition with the maximal value in the weight sum of the weight sum of the corresponding key to the issue word that the match is successful of same user and field keyword, with the numerical value of gained after the addition degree of correlation as this problem to be recommended and this user.Obtain after this problem to be recommended and each user's the degree of correlation, then utilize problem push unit 35 can obtain a plurality of degrees of correlation corresponding with these a plurality of users, these a plurality of degrees of correlation are sorted from high to low, obtain several users corresponding to the degree of correlation that rank in the top, give these several users with the question recommending of recommendation to be with, user's number that this Receiver Problem is recommended can be set according to actual needs.
In embodiments of the present invention, question recommending system in this Ask-Answer Community is by when recommending problem to the user, can be according to field lists of keywords and this matching degree with the recommendation problem in the related field of the tabulation of this user's Weight key to the issue word and this problem of answering with the matching degree of recommendation problem and this user, obtain this problem and user's degree of correlation, thereby with question recommending to the higher user of degree of correlation, so that the user is improved to the response rate of problem in the Ask-Answer Community, the experience effect of user in Ask-Answer Community gets a promotion.
The embodiment of the invention is by mating the key to the issue word tabulation of problem to be recommended with each user's who presets Weight key to the issue word tabulation, obtain the weight sum of the key to the issue word that the match is successful corresponding to this each user, this key to the issue word tabulation is mated with all Weight field lists of keywords of each user who presets respectively, obtain the maximal value in the weight sum of all field keywords that the match is successful corresponding to this each user, and the maximal value in the weight sum of the weight sum of the combination key to the issue word that the match is successful corresponding to same user and field keyword, obtain this problem to be recommended and this user's the degree of correlation, this problem to be recommended is pushed to one or more the highest user of the degree of correlation, solved in the Ask-Answer Community the not high and user's experience effect not good problem of response rate to problem, improve the response rate of user to problem in the Ask-Answer Community, also promoted the experience effect of user in Ask-Answer Community.
The above only is preferred embodiment of the present invention, not in order to limiting the present invention, all any modifications of doing within the spirit and principles in the present invention, is equal to and replaces and improvement etc., all should be included within protection scope of the present invention.
Claims (10)
1. the method for recommending problem in the Ask-Answer Community is characterized in that, described method comprises the steps:
Obtain the key to the issue word tabulation of problem to be recommended;
Described key to the issue word tabulation and each user's who presets Weight key to the issue word tabulation are mated, obtain the weight sum with the corresponding key to the issue word that the match is successful of described each user;
The tabulation of described key to the issue word is mated with all Weight field lists of keywords of default each user respectively, obtain the maximal value in the weight sum of all field keywords that the match is successful corresponding with described each user;
In conjunction with the maximal value in the weight sum of the weight sum of the key to the issue word that the match is successful corresponding to same user and field keyword, obtain described problem to be recommended and described user's the degree of correlation;
According to described problem to be recommended and each user's the degree of correlation, described problem to be recommended is pushed to one or more the highest user of the degree of correlation.
2. the method for claim 1 is characterized in that, before carrying out the described step that described key to the issue word tabulation and each user's who presets Weight key to the issue word tabulation are mated, described method also comprises:
Obtain the problem list that each user once answered and refused to answer, with described problem list according to time sequence, extract occurrence number in the problem list after the ordering greater than 1 key to the issue word, and generate described user's key to the issue word tabulation;
According to the Weight Acquisition relation of default key to the issue word, obtain described user's Weight key to the issue word tabulation.
3. the method for claim 1 is characterized in that, before carrying out the described step that described key to the issue word tabulation is mated with all Weight field lists of keywords of default each user respectively, described method also comprises:
Obtain in all problems of designated field occurrence number greater than the set of the field keyword of default value;
Obtain the first frequency that each keyword in the described set occurs in all problems of described designated field, and obtain the second frequency that described each keyword occurs in all problems of all spectra;
According to described each keyword first frequency and described second frequency, obtain the weight of each keyword of this designated field, and form the Weight field lists of keywords of described designated field.
4. the method for claim 1, it is characterized in that, described the tabulation of described key to the issue word and default each user's Weight key to the issue word tabulation are mated, obtain with the step of the weight sum of the corresponding key to the issue word that the match is successful of described each user and specifically comprise:
The Weight key to the issue word tabulation of a user among the tabulation of the key to the issue word of described problem to be recommended and default all users is mated, obtain and be present in simultaneously the key to the issue word that the match is successful in described lists of keywords and the tabulation of described Weight key to the issue word;
Based on described Weight key to the issue word tabulation, obtain the weight corresponding with the described key to the issue word that the match is successful;
With weight addition corresponding to the described and described key to the issue word that the match is successful, obtain the weight sum with the corresponding key to the issue word that the match is successful of a described user;
Upgrade a described user for the next user among all default users, repeat above-mentioned steps, until obtain with described default all users in the weight sum of the corresponding key to the issue word that the match is successful of each user after, withdraw from.
5. the method for claim 1, it is characterized in that, described the tabulation of described key to the issue word is mated with all Weight field lists of keywords of default each user respectively, the peaked step of obtaining in the weight sum of all field keywords that the match is successful corresponding with described each user specifically comprises:
A. obtain the related field of all problems that a user among all default users answered;
B. according to the Weight field lists of keywords in each field in the default all spectra, obtain all Weight field lists of keywords of the described user corresponding with described related field;
C. a Weight field lists of keywords in all Weight field lists of keywords of the tabulation of the key to the issue word of described problem to be recommended and described user is mated, obtain and be present in simultaneously the field keyword that the match is successful in described key to the issue word tabulation and the described Weight field lists of keywords;
D. based on described each Weight field lists of keywords, obtain the weight sum with the corresponding field keyword that the match is successful of described each Weight field lists of keywords;
E. the next Weight field lists of keywords in all Weight field lists of keywords that to upgrade a Weight field lists of keywords in all Weight field lists of keywords of described user be described user, repeat above-mentioned steps c to d, until obtain with described user the weight sum of the corresponding field keyword that the match is successful of each Weight field lists of keywords after, execution in step f;
F. obtain the maximal value in the described weight sum;
G. upgrade a described user for the next user among all default users, repeat above-mentioned steps a to f, until obtain with described default all users in the weight sum of the corresponding field keyword that the match is successful of each user in maximal value.
6. the method for claim 1 is characterized in that, the weight of described key to the issue word is that scope belongs to the real number of [1,1], the weight of described field keyword be scope belong to (0,1] real number.
7. the question recommending system in the Ask-Answer Community is characterized in that, described system comprises:
Lists of keywords is obtained master unit, is used for obtaining the key to the issue word tabulation of problem to be recommended;
Weight sum acquiring unit is used for described key to the issue word tabulation and each user's who presets Weight key to the issue word tabulation are mated, and obtains the weight sum with the corresponding key to the issue word that the match is successful of described each user;
The maximal value acquiring unit, be used for described key to the issue word tabulation is mated with all Weight field lists of keywords of default each user respectively, obtain the maximal value in the weight sum of all field keywords that the match is successful corresponding with described each user;
Degree of correlation acquiring unit is used for the maximal value in conjunction with the weight sum of the weight sum of the key to the issue word that the match is successful corresponding to same user and field keyword, obtains described problem to be recommended and described user's the degree of correlation; And
The problem push unit is used for the degree of correlation according to described problem to be recommended and each user, and described problem to be recommended is pushed to one or more the highest user of the degree of correlation.
8. system as claimed in claim 7 is characterized in that, described system also comprises:
The tabulation generation unit, be used for obtaining the problem list that each user once answered and refused to answer, with described problem list according to time sequence, extract occurrence number in the problem list after the ordering greater than 1 key to the issue word, and generate described user's key to the issue word tabulation; And
Weight tabulation acquiring unit is used for the Weight Acquisition relation according to default key to the issue word, obtains described user's Weight key to the issue word tabulation.
9. system as claimed in claim 7 is characterized in that, described system also comprises:
The set acquiring unit is used for obtaining all problems occurrence number of designated field greater than the set of the field keyword of default value;
The frequency acquisition unit is used for obtaining the first frequency that each keyword of described set occurs, and obtains the second frequency that described each keyword occurs in all problems of all spectra in all problems of described designated field; And
Tabulation generation unit in Weight field is used for according to described each keyword first frequency and described second frequency, obtains the weight of each keyword of this designated field, and generates the Weight field lists of keywords of described designated field.
10. system as claimed in claim 7 is characterized in that, described weight sum acquiring unit specifically comprises:
The matching keywords acquiring unit, be used for will described problem to be recommended the tabulation of key to the issue word mate with default all users' a user's Weight key to the issue word tabulation, obtain and be present in simultaneously described lists of keywords and the described Weight key to the issue word key to the issue word that the match is successful in tabulating;
The Weight Acquisition unit is used for obtaining the weight corresponding with the described key to the issue word that the match is successful based on described Weight key to the issue word tabulation; And
The weight sum is obtained subelement, is used for weight addition corresponding to the described and described key to the issue word that the match is successful, obtains the weight sum with the corresponding key to the issue word that the match is successful of a described user; And
Updating block, be used for upgrading a described user and be next user of all the default users, trigger described matching keywords acquiring unit, until obtain with described default all users in the weight sum of the corresponding key to the issue word that the match is successful of each user after, withdraw from.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210095917.5A CN103365899B (en) | 2012-04-01 | 2012-04-01 | The problem of in a kind of Ask-Answer Community, recommends method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210095917.5A CN103365899B (en) | 2012-04-01 | 2012-04-01 | The problem of in a kind of Ask-Answer Community, recommends method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103365899A true CN103365899A (en) | 2013-10-23 |
| CN103365899B CN103365899B (en) | 2017-10-20 |
Family
ID=49367263
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210095917.5A Active CN103365899B (en) | 2012-04-01 | 2012-04-01 | The problem of in a kind of Ask-Answer Community, recommends method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103365899B (en) |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103761314A (en) * | 2014-01-26 | 2014-04-30 | 句容云影响软件技术开发有限公司 | Multifunctional conversation information control method |
| 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 |
| CN105205159A (en) * | 2015-09-29 | 2015-12-30 | 努比亚技术有限公司 | Automatic information feedback device and method |
| WO2016101765A1 (en) * | 2014-12-25 | 2016-06-30 | 北京奇虎科技有限公司 | Question-and-answer page related question recommendation method and device |
| CN105912697A (en) * | 2016-04-25 | 2016-08-31 | 北京光年无限科技有限公司 | Optimization method and device of dialog system knowledge base |
| CN106055547A (en) * | 2015-04-02 | 2016-10-26 | 松下知识产权经营株式会社 | Conversation method, recording medium and conversation 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 |
| CN106997383A (en) * | 2017-03-22 | 2017-08-01 | 山东大学 | The problem of based on PageRank algorithms, recommends method and system |
| WO2017202138A1 (en) * | 2016-05-26 | 2017-11-30 | 北京搜狗科技发展有限公司 | Information processing method and apparatus based on search engine |
| CN107451199A (en) * | 2017-07-05 | 2017-12-08 | 阿里巴巴集团控股有限公司 | Method for recommending problem and device, equipment |
| CN107483595A (en) * | 2017-08-23 | 2017-12-15 | 北京百度网讯科技有限公司 | Information-pushing method and device |
| CN107766536A (en) * | 2017-10-30 | 2018-03-06 | 江西博瑞彤芸科技有限公司 | The searching method of related information |
| CN107992604A (en) * | 2017-12-14 | 2018-05-04 | 北京搜狗科技发展有限公司 | The distribution method and relevant apparatus of a kind of task entry |
| CN108090055A (en) * | 2016-11-21 | 2018-05-29 | 凌今 | The help information method for pushing and equipment of a kind of scientific research interaction platform |
| CN109460504A (en) * | 2018-09-21 | 2019-03-12 | 广州神马移动信息科技有限公司 | The answer main body recommended method and its device, electronic equipment, computer-readable medium of answer are reserved in Knowledge Community |
| CN109992602A (en) * | 2019-04-02 | 2019-07-09 | 海南颖川科技有限公司 | Juvenile's digital reading guiding apparatus |
| CN111538826A (en) * | 2020-07-13 | 2020-08-14 | 智者四海（北京）技术有限公司 | Routing invitation method and device based on problems |
| CN112765326A (en) * | 2021-01-27 | 2021-05-07 | 西安电子科技大学 | Question-answering community expert recommendation method, system and application |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6618722B1 (en) * | 2000-07-24 | 2003-09-09 | International Business Machines Corporation | Session-history-based recency-biased natural language document search |
| 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 |
| US20120030126A1 (en) * | 2008-11-26 | 2012-02-02 | Mccagg Brin | Structured Job Search Engine with Guided Navigation |
-
2012
- 2012-04-01 CN CN201210095917.5A patent/CN103365899B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6618722B1 (en) * | 2000-07-24 | 2003-09-09 | International Business Machines Corporation | Session-history-based recency-biased natural language document search |
| 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 |
| US20120030126A1 (en) * | 2008-11-26 | 2012-02-02 | Mccagg Brin | Structured Job Search Engine with Guided Navigation |
Non-Patent Citations (1)
| Title |
|---|
| 呼大为: "面向问答系统的答案获取方法研究与实现", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| 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 |
| CN103761314A (en) * | 2014-01-26 | 2014-04-30 | 句容云影响软件技术开发有限公司 | Multifunctional conversation information control method |
| WO2016101765A1 (en) * | 2014-12-25 | 2016-06-30 | 北京奇虎科技有限公司 | Question-and-answer page related question recommendation method and device |
| CN106055547B (en) * | 2015-04-02 | 2021-06-18 | 松下知识产权经营株式会社 | Dialogue method, recording medium, and dialogue system |
| CN106055547A (en) * | 2015-04-02 | 2016-10-26 | 松下知识产权经营株式会社 | Conversation method, recording medium and conversation 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 |
| CN105205159B (en) * | 2015-09-29 | 2020-06-02 | 陈中和 | Device and method for automatically feeding back information |
| CN105205159A (en) * | 2015-09-29 | 2015-12-30 | 努比亚技术有限公司 | Automatic information feedback device and method |
| CN105912697A (en) * | 2016-04-25 | 2016-08-31 | 北京光年无限科技有限公司 | Optimization method and device of dialog system knowledge base |
| CN105912697B (en) * | 2016-04-25 | 2019-08-27 | 北京光年无限科技有限公司 | A kind of optimization method and device of conversational system knowledge base |
| WO2017202138A1 (en) * | 2016-05-26 | 2017-11-30 | 北京搜狗科技发展有限公司 | Information processing method and apparatus based on search engine |
| CN108090055A (en) * | 2016-11-21 | 2018-05-29 | 凌今 | The help information method for pushing and equipment of a kind of scientific research interaction platform |
| CN106997383A (en) * | 2017-03-22 | 2017-08-01 | 山东大学 | The problem of based on PageRank algorithms, recommends method and system |
| CN107451199B (en) * | 2017-07-05 | 2020-06-26 | 阿里巴巴集团控股有限公司 | Question recommendation method, device and equipment |
| CN107451199A (en) * | 2017-07-05 | 2017-12-08 | 阿里巴巴集团控股有限公司 | Method for recommending problem and device, equipment |
| CN107483595A (en) * | 2017-08-23 | 2017-12-15 | 北京百度网讯科技有限公司 | Information-pushing method and device |
| CN107766536A (en) * | 2017-10-30 | 2018-03-06 | 江西博瑞彤芸科技有限公司 | The searching method of related information |
| CN107992604B (en) * | 2017-12-14 | 2020-08-28 | 北京搜狗科技发展有限公司 | Task item distribution method and related device |
| CN107992604A (en) * | 2017-12-14 | 2018-05-04 | 北京搜狗科技发展有限公司 | The distribution method and relevant apparatus of a kind of task entry |
| CN109460504A (en) * | 2018-09-21 | 2019-03-12 | 广州神马移动信息科技有限公司 | The answer main body recommended method and its device, electronic equipment, computer-readable medium of answer are reserved in Knowledge Community |
| CN109992602A (en) * | 2019-04-02 | 2019-07-09 | 海南颖川科技有限公司 | Juvenile's digital reading guiding apparatus |
| CN109992602B (en) * | 2019-04-02 | 2023-05-16 | 海南颖川科技有限公司 | Digital reading guiding equipment for children |
| CN111538826B (en) * | 2020-07-13 | 2021-01-29 | 智者四海（北京）技术有限公司 | Routing invitation method and device based on problems |
| CN111538826A (en) * | 2020-07-13 | 2020-08-14 | 智者四海（北京）技术有限公司 | Routing invitation method and device based on problems |
| CN112765326A (en) * | 2021-01-27 | 2021-05-07 | 西安电子科技大学 | Question-answering community expert recommendation method, system and application |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103365899B (en) | 2017-10-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103365899A (en) | Question recommending method and question recommending system both in questions-and-answers community | |
| CN108829822B (en) | Media content recommendation method and device, storage medium and electronic device | |
| CN102103634B (en) | Information processing apparatus and method | |
| JP6677419B2 (en) | Voice interaction method and apparatus | |
| US20200304433A1 (en) | Interactive virtual conversation interface systems and methods | |
| CN109543007A (en) | Put question to data creation method, device, computer equipment and storage medium | |
| CN108846138B (en) | Question classification model construction method, device and medium fusing answer information | |
| CN103491205A (en) | Related resource address push method and device based on video retrieval | |
| CN111882370B (en) | Advertisement recommendation method and device and electronic equipment | |
| CN104978314A (en) | Media content recommendation method and device | |
| CN113672708A (en) | Language model training method, question and answer pair generation method, device and equipment | |
| CN116881428B (en) | Language model training method and device | |
| CN112650842A (en) | Human-computer interaction based customer service robot intention recognition method and related equipment | |
| KR20200014047A (en) | Method, system and computer program for knowledge extension based on triple-semantic | |
| CN110895656A (en) | Text similarity calculation method and device, electronic equipment and storage medium | |
| CN117251552B (en) | Dialogue processing method and device based on large language model and electronic equipment | |
| CN116541493A (en) | Interactive response method, device, equipment and storage medium based on intention recognition | |
| CN116738250A (en) | Prompt text expansion method, device, electronic equipment and storage medium | |
| CN116150306A (en) | Training method of question-answering robot, question-answering method and device | |
| CN117271736A (en) | Question-answer pair generation method and system, electronic equipment and storage medium | |
| CN114090766A (en) | Video text screening method and device and electronic equipment | |
| CN109857843A (en) | Exchange method and system based on document | |
| CN117421413A (en) | Question-answer pair generation method and device and electronic equipment | |
| CN112307754A (en) | Statement acquisition method and device | |
| CN105653619A (en) | Update method and device of correct log library in intelligent question-answering system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| ASS | Succession or assignment of patent right |
Owner name: SHENZHEN SHIJI LIGHT SPEED INFORMATION TECHNOLOGY Free format text: FORMER OWNER: TENGXUN SCI-TECH (SHENZHEN) CO., LTD. Effective date: 20131029 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 518044 SHENZHEN, GUANGDONG PROVINCE TO: 518057 SHENZHEN, GUANGDONG PROVINCE |
|
| TA01 | Transfer of patent application right |
Effective date of registration: 20131029 Address after: A Tencent Building in Shenzhen Nanshan District City, Guangdong streets in Guangdong province science and technology 518057 16 Applicant after: Shenzhen Shiji Guangsu Information Technology Co., Ltd. Address before: Shenzhen Futian District City, Guangdong province 518044 Zhenxing Road, SEG Science Park 2 East Room 403 Applicant before: Tencent Technology (Shenzhen) Co., Ltd. |
|
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |