CN105447080B - A kind of inquiry complementing method in community's question and answer search - Google Patents
A kind of inquiry complementing method in community's question and answer search Download PDFInfo
- Publication number
- CN105447080B CN105447080B CN201510745059.8A CN201510745059A CN105447080B CN 105447080 B CN105447080 B CN 105447080B CN 201510745059 A CN201510745059 A CN 201510745059A CN 105447080 B CN105447080 B CN 105447080B
- Authority
- CN
- China
- Prior art keywords
- question
- word
- user
- inquiry
- answer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/242—Query formulation
- G06F16/2425—Iterative querying; Query formulation based on the results of a preceding query
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2457—Query processing with adaptation to user needs
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A kind of inquiry complementing method in being searched for the present invention relates to community's question and answer, belongs to technical field of information retrieval；Include the following steps：One, cQA databases are built；Two, entity dictionary and phrase dictionary are built；Three, for query statement input by user, judge whether the last one word is complete, if imperfect, carry out end word completion；Four, query statement is ranked up the problems in database by using ranking functions to obtain initial candidate problem list；Five, list is required to screen according to diversity, size appropriateness, local isotonicity and fidelity；Six, it is sorted again according to the importance and quality of candidate problem；Seven, the preceding N items candidate problem after output is sorted again is selected for user.Compared with prior art, the present invention can carry out inquiry completion when lacking user and searching for daily record, overcome recommend Similar Problems in existing method, the problem of size is not inconsistent, relative ranks are constant, distortion, be better than the using effect of existing search engine on user experience.
Description
Technical field
The present invention relates to a kind of inquiry complementing method more particularly to it is a kind of can community's question and answer search in inquiry completion
Method, the user in capable of being searched for community's question and answer, which inquires, carries out auto-complete, effectively promotes user experience, belongs to information retrieval
Technical field.
Background technology
With the development of internet, the informationization of human society, networking process are greatly speeded up, and information retrieval replaces already
Network times are stepped into manual information retrieval.In information retrieval field, inquiry completion technology is for user's search and expressing information demand
It is very helpful.User usually inputs a brief query statement rather than a complete problem in retrieval, and right
In retrieval model, the information that a complete problem is provided will be far superior to a brief query statement.Therefore, work as user
When input inquiry sentence, the query statement that user provides complete problem form how to be helped to have great importance.
Currently, inquiry completion technology has been achieved for certain progress.They mainly excavate user and search for daily record and net
Page content information, the search engine of some mainstreams additionally provide the recommendation of related phrase.Community-based question and answer are searched for
(cQA), it carries out inquiring the work that completion is a very challenging property in the case where lacking user's search daily record.Currently, community
Inquiry completion technology in question and answer search is still at an early stage.
Invention content
The purpose of the present invention is be directed to how lack user search for daily record when community's question and answer search in carry out inquiry benefit
A kind of full problem, it is proposed that inquiry complementing method based on sequence.This method can provide several complete times to the user
Select problem selective, it is effective that user is helped to carry out inquiry completion, greatly improve user experience.
To achieve the above object, the technical solution adopted in the present invention is as follows：
First with community's question and answer of crawl to building cQA databases, web page title on wikipedia and common is utilized
Phrase dictionary creation entity and phrase database；Then judge whether the last one word of query statement input by user is complete, if
End word is imperfect, utilizes correlation rule completion end word；Finally, for complete query statement, to candidate problem into
The initial sequence of row is screened and is sorted again.
The specific technical solution of the present invention is as follows：
Inquiry complementing method in a kind of search of community's question and answer, this approach includes the following steps：
Step 1: based on question and answer language material structure by question and answer to the cQA databases that form；
Step 2: based on existing encyclopaedic knowledge resource and dictionary resources structure entity dictionary and phrase dictionary；
Step 3: for query statement input by user, judge whether the last one word of query statement is complete, if last
Tail word is imperfect, goes to step four；Otherwise, five are gone to step；
Step 4: carrying out end word completion to the query statement of input obtains complete query statement；
Step 5: being ranked up by using ranking functions to obtain initial confession to the problems in database to query statement
The candidate problem list of user's selection；
Preferably, in order to improve retrieval effectiveness, the smooth method of linear interpolation is used in ranking functions and statistics is turned over
The model translated, it is specific as follows：
p(w|(q, a))=(1- λ) pmx(w|(q, a))+λ pml(w|C) (2)
Wherein, q is the query statement of user, and w is each word in query statement, C=(q, a)1, (q, a
)2..., (q, a)LIndicate cQA databases in question and answer pair, (q, a)iIt is i-th of question and answer pair, q problem of representation, a expression answers；
λ is smoothing parameter, pml(w|C it) indicates to give C, the conditional probability of word w；pml(w|Q) it indicates to give problem q, the condition of word w is general
Rate；p(w|T) it indicates to give the word t in problem, the conditional probability of word w；pml(t|Q) it indicates to give problem q, the condition of word t
Probability；pml(w|A) it indicates to give answer a, the conditional probability of word w；And meet alpha+beta+γ=1.
Step 6: the initial candidate problem list that step 4 is obtained according to diversity, size appropriateness, local isotonicity and
The requirement of fidelity is screened；
Pass through preferably, the requirement according to diversity, size appropriateness, local isotonicity and fidelity carries out screening
Following methods are realized：
(1) in order to meet the requirement of diversity, the method using topic model (topic model, LDA) is each candidate
Problem distributes theme, and the candidate problem for limiting return should come from different themes；
(2) in order to meet the requirement of size appropriateness, the length for returning to candidate problem is limited no more than some threshold value s；
(3) in order to meet the requirement of local isotonicity, limiting the relative ranks of query terms can change, but query statement
In the partial order of entity and phrase that is identified based on entity dictionary and phrase dictionary cannot change；
(4) in order to meet the requirement of fidelity, it must includes word input by user to limit in candidate problem；
Preferably, the threshold value s is 62 characters.
Preferably, the entity and phrase that are identified based on entity dictionary and phrase dictionary in the query statement are using most
The method of big String matching is identified.
Step 7: to the candidate problem list after screening according to importance and the quality sequence from high to low of candidate problem
It is sorted again；
Preferably, technorati authority and user degree of attentiveness characterization of the importance using question and answer pair in cQA databases.
Preferably, the value of the importance after technorati authority and user's degree of attentiveness regularization by equal weight by will be added
It is calculated together.
Preferably, the quality is characterized using the spelling in text, syntax error.
Preferably, the value of the quality is calculated by the counting how many times occurred to mistake.
Preferably, further, in order to help user quickly identify output candidate problem difference, to each candidate
The very corn of a subject word carries out overstriking processing, does delete processing to invalid title, to promote user experience.
Step 8: the preceding N items candidate problem after output is sorted again is selected for user.
Advantageous effect
Compared with common inquiry complementing method, the present invention can be searched for when lacking user and searching for daily record in community's question and answer
In carry out inquiry completion, using the thought recommended candidate problem based on sequence, meet every demand in cQA search, overcome
In traditional directory complementing method recommend Similar Problems, size is not inconsistent, relative ranks are constant, distortion the problem of.The present invention can have
Effect ground helps user to carry out inquiry completion, is better than existing search engine using effect on user experience, meanwhile, energy of the present invention
It is enough flexibly to be combined to obtain better effect with existing inquiry complementing method.
Description of the drawings
Fig. 1 is the inquiry complementing method flow diagram in a kind of community's question and answer search of the embodiment of the present invention.
Fig. 2 is inquiry completion exemplary graph (a：Effect is poor；b：Effect is good).
Fig. 3 is inquiry completion comparison result (a：Yahoo；b：Google；c：Method proposed by the present invention)
Specific implementation mode
With reference to embodiment and attached drawing, the specific implementation mode of the present invention is described in further detail.
Inquiry complementing method in a kind of community's question and answer search of the present invention, as shown in Figure 1, including the following steps：
One, it is built by question and answer to the cQA databases that form based on question and answer language material.
We are from Yahoo！Language material (problem and corresponding answer) is crawled on Answer, is constituted one and is surrounded by 6,345,
The cQA databases of 786 question and answer pair.Based on this database, we have developed the systems of cQA search, provide and search
The function of rope and inquiry completion.
Two, it is based on existing encyclopaedic knowledge resource (such as Baidupedia or wikipedia) and dictionary resources builds entity word
Allusion quotation and phrase dictionary.
In order to identify the entity in user's query statement, Dictionary use maximum string that we are made of by one entity
The method matched identifies.In the present embodiment because the page title of wikipedia can with presentation-entity name, we use
Wikipedia web page title builds entity dictionary.Identification for phrase, we use same method, that is, utilize Dictionary use
The method of maximum String matching；And for the dictionary of phrase, we utilize the dictionary for commonly using English phrase.
Three, for query statement input by user, judge whether the last one word of query statement is complete, if end is single
Word is imperfect, goes to step four；Otherwise, five are gone to step.
Four, end word completion is carried out to the query statement of input and obtains complete query statement.
The method of completion end word realizes that steps are as follows：
In default of search daily record, we use famous correlation rule learning algorithm (such as Apriori of Data Mining
Algorithm) in language material acquistion is gone to school to correlation rule, the correlation rule and its frequency of occurrences that these study obtain are similar to search skill
Search daily record in art, to be realized using the word completion technology in search technique.
Five, query statement is ranked up to obtain by using ranking functions to the problems in database initial for user
The candidate problem list of selection.
In general search, the frequency of occurrences for searching for query statement in daily record is a key factor of inquiry completion, but
It is the method that we use ranked candidate problem due to lacking search daily record.Here, it is contemplated that due to much having difference
The problem of semantic structure and the form of expression, all expresses equivalent, for example, problem " you be able to tell that my iphone5 and
The difference of iphone5s" and " what the difference of iphone5 and iphone5s is", the two has different syntactic structures but
The identical meaning is expressed, therefore cannot be sorted using the frequency of occurrences of problem.It is initial in order to obtain one in view of these
Sorted lists, we are ranked up the problems in database using a ranking functions.Formula is as follows：
p(w|(q, a))=(1- λ) pmx(w|(q, a))+λ pml(w|C) (2)
Wherein, q is the query statement of user, and w is that each word in query statement (for Chinese, needs in first doing
Text participle；For English, using space-separated words), C=(q, a)1, (q, a)2..., (q, a)LIndicate in cQA databases
Question and answer pair, (q, a)iIt is i-th of question and answer pair, q problem of representation, a expression answers；λ is smoothing parameter, pml(w|C) indicate given
C, the conditional probability of word w；pml(w|Q) it indicates to give problem q, the conditional probability of word w；p(w|T) it indicates to give the word in problem
T, the conditional probability of word w；pml(t|Q) it indicates to give problem q, the conditional probability of word t；pml(w|A) it indicates to give answer a, it is single
The conditional probability of word w；And meet alpha+beta+γ=1.
By the way that with superior function, we can calculate the problems in cQA databases to the possibility of generation query statement, borrow
This is ranked up problem, generates initial candidate problem list.
It is necessary to explanation, in the present embodiment, we use semantic relation of the statistical translation model between word and word
It is portrayed, while linear interpolation smoothing method having been used to be combined background language model and the advantages of translation model,
To be conducive to improve sequence effect.
Six, the initial candidate problem list obtained to step 4 is according to diversity, size appropriateness, local isotonicity and fidelity
The requirement of property is screened.
Inquiry completion technology in cQA search needs to meet following four demand：
(1) diversity：As shown in Fig. 2 (a), user inputs " iphone 3gs iphone 3g ", and click " search " is pressed afterwards
Button, what system provided recommends problem mostly to express the similar meaning as shown in drop-down list, this can waste user's browsing
Time, reduce user experience effect, and Fig. 2 (b) have preferable diversity；
(2) size appropriateness：As shown in Fig. 2 (a), the problem of the last one recommendation, has exceeded the size of Suggestion box, so not
The problem of only influence is beautiful, while reducing user experience effect, and recommended in Fig. 2 (b) size appropriateness；
(3) local isotonicity：For entity and phrase, their partial order is kept to be necessary, without necessity
The relative ranks between query terms are kept in candidate problem；
(4) fidelity：Inquiry completion technology cannot lose the primitive meaning of query statement input by user.
For needing above, we screen initial sorted lists：
(1) in order to meet the requirement of diversity, we are each using the method for topic model (topic model, LDA)
Candidate problem distributes theme, and the candidate problem for limiting return should come from different themes；
(2) in order to meet the requirement of size appropriateness, we limit the length for returning to candidate problem no more than some threshold
Value, here, the threshold value that we set is 62 characters；
(3) in order to meet the requirement of local isotonicity, the relative ranks that we limit query terms can change, but inquire
The partial order of the entity and phrase that are identified based on entity dictionary and phrase dictionary in sentence cannot change；
(4) in order to meet the requirement of fidelity, it must include word input by user that we, which limit in candidate problem,.
Seven, the candidate problem list after screening is carried out according to importance and the quality sequence from high to low of candidate problem
It sorts again.
After initially sorting and screening, we can only obtain a coarse problem sorted lists.Sequencer procedure again
In we the importance and quality of candidate problem are assessed.
Importance：We weigh the importance of problem using the technorati authority of question and answer pair and user's degree of attentiveness.(1) one
The quantity for the best problem that user answers can be used for weighing the technorati authority of the user, and therefore, we use asker and answer
Technorati authority of the technorati authority summation of person as the question and answer pair.(2) question and answer can indicate the whole quantity of answering for being included
Degree of attentiveness of the user to the problem.To sum up, we will be added together problem of representation after this two regularizations by equal weight
Importance.
Quality：CQA data be issued by user, and user publication text generally existing some spelling, syntax errors
Deng.Therefore, we weigh the quality of problem using the number that mistake occurs.
We sort again to the list after screening after considering the importance and quality of problem, meanwhile, in order to help
User quickly identifies the difference of different candidate problems, we are to each candidate the very corn of a subject word (such as interrogative, verb and name
Word) overstriking processing has been carried out, the user experience is improved in this way.As shown in Fig. 3 (c), for the core word in candidate problem, we
Overstriking processing has been carried out so that it is convenient to which user identifies.In addition, for invalid title, we do delete processing.
Eight, the preceding N items candidate problem after output is sorted again is selected for user.
Test result
Effective in order to verify the inquiry complementing method in cQA search provided in this embodiment, we sample from data set
25 problems select the query statement for building the problem containing 2,3,4,5 keywords from these problems, share 100 in this way
A query statement.For example, for problem " what is the difference between iphone 5and iphone
5s", its corresponding query statement be " iphone 5 ", " iphone 5s ", " 5 5s difference of iphone ",
"iphone 5 5s what difference".For each problem, we use its respective queries sentence checking system
The problem of in the recommendation list of return whether comprising the problem or with identical meaning, if including, mark its sequence position
It sets.The results are shown in table below for recall rate：
The recall rate (%) of 1 present invention of table
| Keyword number contained by query statement | Preceding 1 | Preceding 2 | Preceding 6 | Preceding 10 |
| 2 | 20 | 28 | 44 | 72 |
| 3 | 52 | 68 | 80 | 80 |
| 4 | 72 | 80 | 84 | 84 |
| 5 | 76 | 80 | 84 | 84 |
The experimental results showed that when inputting 4 keywords, preceding 2 recommendation problems, which are the desired possibilities of user, is
80%.This demonstrate that the present invention proposes the validity of method.
In fig. 3, we using method proposed by the present invention and two comparison mainstreams search engine (Google and
Yahoo it) is compared, Fig. 3 (a) and 3 (b) maintain the relative ranks between query terms, and the question sentence of completion is not unobstructed enough
Fluently, and user often input when it is more random, will not emphasize the relative ranks of word, this just need inquire completion system energy handle
Query terms are organized into more clear and more coherent question sentence.Fig. 3 (c) is method proposed by the present invention, as a result it can be seen that the method for the present invention
Better than them.The two search engines possess a large amount of search daily record and Internet resources, and therefore, we are contemplated that the present invention's
For method in the case where possessing same asset, the effect for inquiring completion is centainly better than existing search engine.
The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention.The technology of the industry
Personnel are it should be appreciated that the present invention is not limited to the above embodiments, and the above embodiments and description only describe this
The principle of invention, without departing from the spirit and scope of the present invention, various changes and improvements may be made to the invention, these changes
Change and improve all within the scope of the claimed invention, the claimed scope of the invention is by appended claims and its waits
Effect object defines.
Claims (9)
1. the inquiry complementing method in a kind of community's question and answer search, it is characterised in that：
Step 1: based on question and answer language material structure by question and answer to the cQA databases that form；
Step 2: based on existing encyclopaedic knowledge resource and dictionary resources structure entity dictionary and phrase dictionary；
Step 3: for query statement input by user, judge whether the last one word of query statement is complete, if end is single
Word is imperfect, goes to step four；Otherwise, five are gone to step；
Step 4: carrying out end word completion to the query statement of input obtains complete query statement；
Step 5: by using ranking functions being ranked up to obtain to the problems in database to query statement initial for user
The candidate problem list of selection；
Step 6: the initial candidate problem list obtained to step 4 is according to diversity, size appropriateness, local isotonicity and fidelity
The requirement of property is screened；
Step 7: being carried out according to importance and the quality sequence from high to low of candidate problem to the candidate problem list after screening
It sorts again；
Step 8: the preceding N items candidate problem after output is sorted again is selected for user；
In order to improve retrieval effectiveness, the smooth method of linear interpolation and statistical translation are used in ranking functions described in step 5
Model, it is specific as follows：
p(w|(q, a))=(1- λ) pmx(w|(q, a))+λ pml(w|C) (2)
Wherein, q is the query statement of user, and w is each word in query statement, C=(q, a)1, (q, a)2..., (q,
a)LIndicate cQA databases in question and answer pair, (q, a)iIt is i-th of question and answer pair, q problem of representation, a expression answers；λ is smoothly to join
Number, pml(w|C it) indicates to give C, the conditional probability of word w；pml(w|Q) it indicates to give problem q, the conditional probability of word w；p(w|t)
It indicates to give the word t in problem, the conditional probability of word w；pml(t|Q) it indicates to give problem q, the conditional probability of word t；pml
(w|A) it indicates to give answer a, the conditional probability of word w；And meet alpha+beta+γ=1.
2. the inquiry complementing method in a kind of community's question and answer search according to claim 1, which is characterized in that the basis
The requirement of diversity, size appropriateness, local isotonicity and fidelity carries out screening and is realized by the following method：
(1) in order to meet the requirement of diversity, the method using topic model (topic model, LDA) is each candidate problem
Theme is distributed, and the candidate problem for limiting return should come from different themes；
(2) in order to meet the requirement of size appropriateness, the length for returning to candidate problem is limited no more than some threshold value s；
(3) in order to meet the requirement of local isotonicity, limiting the relative ranks of query terms can change, but base in query statement
It cannot change in the partial order of entity and phrase that entity dictionary and phrase dictionary identify；
(4) in order to meet the requirement of fidelity, it must includes word input by user to limit in candidate problem.
3. the inquiry complementing method in a kind of community's question and answer search according to claim 2, which is characterized in that the threshold value
S is 62 characters.
4. the inquiry complementing method in a kind of community's question and answer search according to claim 2, which is characterized in that the inquiry
The method of the entity and the maximum String matching of phrase use that are identified based on entity dictionary and phrase dictionary in sentence is identified.
5. the inquiry complementing method in a kind of community's question and answer search according to claim 1, which is characterized in that step 7 institute
State technorati authority and user degree of attentiveness characterization of the importance using question and answer pair in cQA databases.
6. the inquiry complementing method in a kind of community's question and answer search according to claim 5, which is characterized in that described important
The value of property is calculated by will be added together by equal weight after technorati authority and user's degree of attentiveness regularization.
7. the inquiry complementing method in a kind of community's question and answer search according to claim 5, which is characterized in that the quality
It is characterized using the spelling in text, syntax error.
8. the inquiry complementing method in a kind of community's question and answer search according to claim 7, which is characterized in that the quality
Value by mistake occur counting how many times calculate.
9. the inquiry complementing method in a kind of community's question and answer search according to claim 1, which is characterized in that in order to help
User quickly identifies the difference of the candidate problem of output, is also needed to after the completion of the step 7 to each candidate the very corn of a subject
Word carries out overstriking processing, does delete processing to invalid title, to promote user experience.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510745059.8A CN105447080B (en) | 2015-11-05 | 2015-11-05 | A kind of inquiry complementing method in community's question and answer search |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510745059.8A CN105447080B (en) | 2015-11-05 | 2015-11-05 | A kind of inquiry complementing method in community's question and answer search |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105447080A CN105447080A (en) | 2016-03-30 |
| CN105447080B true CN105447080B (en) | 2018-10-26 |
Family
ID=55557258
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510745059.8A Active CN105447080B (en) | 2015-11-05 | 2015-11-05 | A kind of inquiry complementing method in community's question and answer search |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105447080B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107092609B (en) * | 2016-05-10 | 2021-04-02 | 口碑控股有限公司 | Information pushing method and device |
| CN106294656B (en) * | 2016-08-04 | 2019-03-19 | 武汉大学 | A kind of method of map locating keyword to relevant issues |
| CN107016561B (en) * | 2016-10-28 | 2020-10-20 | 创新先进技术有限公司 | Information processing method and device |
| CN108074116B (en) * | 2016-11-09 | 2022-02-22 | 阿里巴巴集团控股有限公司 | Information providing method and device |
| CN108509476A (en) * | 2017-09-30 | 2018-09-07 | 平安科技（深圳）有限公司 | Problem associates method for pushing, electronic device and computer readable storage medium |
| CN109828981B (en) * | 2017-11-22 | 2023-05-23 | 阿里巴巴集团控股有限公司 | Data processing method and computing device |
| CN108763251B (en) * | 2018-04-02 | 2021-06-01 | 创新先进技术有限公司 | Personalized recommendation method and device for nuclear product and electronic equipment |
| CN110750704B (en) * | 2019-10-23 | 2022-03-11 | 深圳计算科学研究院 | Method and device for automatically completing query |
| CN111209382B (en) * | 2020-01-03 | 2024-02-27 | 联想(北京)有限公司 | Content processing method, device, electronic equipment and medium |
| CN116955577B (en) * | 2023-09-21 | 2023-12-15 | 四川中电启明星信息技术有限公司 | Intelligent question-answering system based on content retrieval |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2181406A1 (en) * | 2007-07-11 | 2010-05-05 | Koninklijke Philips Electronics N.V. | Method of operating an information retrieval system |
| CN102073736A (en) * | 2011-01-20 | 2011-05-25 | 百度在线网络技术（北京）有限公司 | Method and system for searching difficult word |
| CN104123322A (en) * | 2013-04-28 | 2014-10-29 | 百度在线网络技术（北京）有限公司 | Method and device for obtaining related question corresponding to input question based on synonymy processing |
-
2015
- 2015-11-05 CN CN201510745059.8A patent/CN105447080B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2181406A1 (en) * | 2007-07-11 | 2010-05-05 | Koninklijke Philips Electronics N.V. | Method of operating an information retrieval system |
| CN102073736A (en) * | 2011-01-20 | 2011-05-25 | 百度在线网络技术（北京）有限公司 | Method and system for searching difficult word |
| CN104123322A (en) * | 2013-04-28 | 2014-10-29 | 百度在线网络技术（北京）有限公司 | Method and device for obtaining related question corresponding to input question based on synonymy processing |
Non-Patent Citations (1)
| Title |
|---|
| "专业搜索引擎的无日志查询推荐机制研究及实现";田宇辰;《中国优秀硕士学位论文全文数据库 信息科技辑》;20150115;论文第20-23页第3.1节，图3.2 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105447080A (en) | 2016-03-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105447080B (en) | A kind of inquiry complementing method in community's question and answer search | |
| Brickley et al. | Google Dataset Search: Building a search engine for datasets in an open Web ecosystem | |
| CN106844658B (en) | Automatic construction method and system of Chinese text knowledge graph | |
| KR101661198B1 (en) | Method and system for searching by using natural language query | |
| CN105528437B (en) | A kind of question answering system construction method extracted based on structured text knowledge | |
| Alexander et al. | Natural language web interface for database (NLWIDB) | |
| CN105786991A (en) | Chinese emotion new word recognition method and system in combination with user emotion expression ways | |
| US20120233160A1 (en) | System and method for assisting a user to identify the contexts of search results | |
| CN103440314A (en) | Semantic retrieval method based on Ontology | |
| Feng et al. | Patent text mining and informetric-based patent technology morphological analysis: an empirical study | |
| Van Hooland et al. | Evaluating the success of vocabulary reconciliation for cultural heritage collections | |
| WO2014210387A2 (en) | Concept extraction | |
| Betti et al. | Expert concept-modeling ground truth construction for word embeddings evaluation in concept-focused domains | |
| Souza et al. | Characterising text mining: a systematic mapping review of the portuguese language | |
| Maciołek et al. | Cluo: Web-scale text mining system for open source intelligence purposes | |
| Carrion et al. | A taxonomy generation tool for semantic visual analysis of large corpus of documents | |
| Song et al. | Semi-automatic construction of a named entity dictionary for entity-based sentiment analysis in social media | |
| Xiao et al. | A practice of tourism knowledge graph construction based on heterogeneous information | |
| Kadry et al. | On the improvement of weighted page content rank | |
| Marx et al. | Exploring term networks for semantic search over RDF knowledge graphs | |
| Aslam et al. | SPedia: a central hub for the linked open data of scientific publications | |
| Carmel et al. | Entity oriented search and exploration for cultural heritage collections: the EU cultura project | |
| Wasim et al. | Extracting and modeling user interests based on social media | |
| Goyal et al. | Concept based query recommendation | |
| KR102454261B1 (en) | Collaborative partner recommendation system and method based on user information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |