WO2016101727A1 - 基于问答的搜索结果调整方法和装置 - Google Patents

基于问答的搜索结果调整方法和装置 Download PDF

Info

Publication number
WO2016101727A1
WO2016101727A1 PCT/CN2015/094619 CN2015094619W WO2016101727A1 WO 2016101727 A1 WO2016101727 A1 WO 2016101727A1 CN 2015094619 W CN2015094619 W CN 2015094619W WO 2016101727 A1 WO2016101727 A1 WO 2016101727A1
Authority
WO
WIPO (PCT)
Prior art keywords
question
answer
search
result
knowledge base
Prior art date
Application number
PCT/CN2015/094619
Other languages
English (en)
French (fr)
Inventor
孙林
王建刚
陈培军
Original Assignee
北京奇虎科技有限公司
奇智软件（北京）有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Priority claimed from CN201410816126.6A external-priority patent/CN105786873B/zh
Priority claimed from CN201410812104.2A external-priority patent/CN105786852B/zh
Application filed by 北京奇虎科技有限公司, 奇智软件（北京）有限公司 filed Critical 北京奇虎科技有限公司
Publication of WO2016101727A1 publication Critical patent/WO2016101727A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor

Definitions

  • the invention relates to the field of information search, in particular to a method and device for adjusting search results based on question and answer, and a method and device for integrating search results.
  • the search engine refers to collecting information from the Internet according to a certain strategy and using a specific computer program. After organizing and processing the information, it provides a search service for the user, which will be related to the user search. The information is presented to the user's system.
  • the search results given by the existing search engine are sorted by the relevance of the search words, and the result is disorderly and unintuitive, and the user is difficult to select and has low efficiency. It is also very inconvenient.
  • the present invention has been made in order to provide a Q&A-based search result adjustment method and apparatus that overcomes the above problems or at least partially solves the above problems, and a search result integration method and apparatus.
  • a Q&A-based search result adjustment method including: receiving a search term from a user, analyzing whether the search term is a question-and-answer type search term; and when the search term is a question-and-answer search term And matching the question and answer class result corresponding to the search term in the question and answer knowledge base; adjusting the search result by using the question and answer class result, and providing the adjusted search result to the user.
  • a Q&A-based search result adjusting apparatus comprising: a data input device adapted to receive a search term from a user; and an analyzer adapted to analyze whether the search term is a question and answer class a search term; a matcher adapted to match a question-and-answer class result corresponding to the search term in the question-and-answer knowledge base when the search term is a question-and-answer class search term; and an adjuster adapted to use the question-answer class result to search result Making adjustments; a data exporter adapted to provide the adjusted search results to the user.
  • the invention provides a question and answer based search result adjustment scheme.
  • the search word from the user is a question and answer type search word
  • the question and answer class result corresponding to the search word is matched in the question and answer knowledge base, and then the search result is performed by using the question and answer class result.
  • the question and answer knowledge base can provide specialized knowledge and information related to the question and answer class, the information provided is large, rich and targeted.
  • the Q&A knowledge base matches the question and answer class results corresponding to the search term, and uses the question and answer class result to search.
  • the result is adjusted to provide the adjusted search result to the user, so that the search result provided to the user is more accurate and comprehensive, so that the search result in the related technology is sorted by the relevance of the search word, and the result is disorderly and unintuitive, and the user selects the result. Difficult, less efficient problems.
  • the present invention can automatically analyze the search term. If the search term is a question-and-answer search term, the user directly enters the professional Q&A knowledge base to perform a search to achieve a better understanding of the user's query requirements. Purpose to improve the user's search experience.
  • the present invention provides search results including the results of the question and answer class to the user. It allows users to get the answers or related knowledge of the questions in the most intuitive and convenient way, which simplifies the user's self-screening of messy search results.
  • a search result integration method comprising: receiving a question and answer type search word from a user, searching for a plurality of search items matching the question and answer type search word; extracting the plurality of pieces Search for multiple Q&A class entries in an entry; merge multiple Q&A class entries from the same Q&A knowledge base into one integrated Q&A result.
  • a search result integration apparatus comprising: a data inputter adapted to receive a question and answer type search term from a user; a searcher adapted to search for a match with the question and answer type search term a plurality of search items; an extractor adapted to extract a plurality of question and answer class entries of the plurality of search terms; an integrator adapted to merge the plurality of question and answer class entries from the same question and answer knowledge base into one integrated question and answer result.
  • the search result integration method receives a question and answer type search word from a user, searches for a plurality of search items matching the question and answer type search word, and then extracts multiple question and answer class items in the plurality of search items, which will be from the same question and answer.
  • Multiple Q&A entries in the Knowledge Base are combined into one integrated Q&A result. Therefore, the present invention can understand the user's search intention, that is, can identify the question and answer type search words input by the user, and then extract the question and answer class items from the search items matching the question and answer type search words, thereby realizing the purpose of active screening and improving the user search experience. .
  • the present invention can flexibly and intelligently combine multiple question and answer class entries from the same Q&A knowledge base according to the source of the data resource, and realize classification according to different Q&A knowledge bases, and can be provided by the class when provided to the user. Further, in this class, the user is provided with more and more detailed question and answer class entries, and the amount of information provided to the user is large.
  • multiple question and answer class entries before the merge are displayed on the search result page in the form of one or more integrated question and answer results, thereby saving the location resources of the search result page. Further, the saved location is able to display other search results, so additional information can be provided to the user in the same search results page for additional help.
  • a computer program comprising computer readable code, when said computer readable code is run on a computing device, causing said computing device to perform a question and answer according to said The search result adjustment method and the search result integration method.
  • a computer readable medium storing the above computer program is provided.
  • FIG. 1 shows a flow chart of a question-and-answer based search result adjustment method in accordance with one embodiment of the present invention
  • FIG. 2 is a schematic diagram showing search results of a search term "the practice of tomato eggs” in the related art
  • FIG. 3 is a schematic diagram showing a search result of a search term “the practice of tomato eggs” provided by the technical solution of the present invention
  • FIG. 4 is a schematic diagram showing another search result of the search term “the practice of tomato eggs” provided by the technical solution of the present invention.
  • FIG. 5 is a schematic diagram showing the result of jumping to the Q&A knowledge base to initiate a search request according to the search term when the link element "More related questions" is clicked;
  • FIG. 6 is a schematic structural diagram of a Q&A-based search result adjusting apparatus according to an embodiment of the present invention.
  • FIG. 7 is a schematic structural diagram of a Q&A-based search result adjusting apparatus according to another embodiment of the present invention.
  • FIG. 8 is a flow chart showing a method for integrating search results according to an embodiment of the present invention.
  • FIG. 9 is a schematic diagram showing search results of the question-and-answer search term "the practice of mashed potatoes" provided by the technical solution of the present invention.
  • Figure 10 is a diagram showing the result of jumping to the Q&A knowledge base to initiate a search request according to the search term when clicking the link element "More Related Questions";
  • FIG. 11 is a schematic diagram showing search results of the question and answer type search word provided by the technical solution of the present invention as “nutritive value of eggs”;
  • FIG. 12 is a block diagram showing the structure of a search result integration apparatus according to an embodiment of the present invention.
  • FIG. 13 is a schematic structural diagram of a search result integration apparatus according to another embodiment of the present invention.
  • FIG. 14 is a block diagram schematically showing a computing device for performing a question-answer based search result adjustment method and a search result integration method according to the present invention
  • Fig. 15 schematically shows a storage unit for holding or carrying program code implementing the Q&A-based search result adjustment method and the search result integration method according to the present invention.
  • an embodiment of the present invention provides a method for adjusting search results based on question and answer.
  • 1 shows a flow chart of a Q&A based search result adjustment method in accordance with one embodiment of the present invention. As shown in FIG. 1, the method includes at least the following steps S102 to S106.
  • Step S102 Receive a search term from a user, and analyze whether the search term is a question-and-answer search term.
  • Step S104 When the search term is a question-and-answer type search word, the question-and-answer class result corresponding to the search word is matched in the question-and-answer knowledge base.
  • Step S106 Adjust the search result by using the result of the question and answer class, and provide the adjusted search result to the user.
  • the invention provides a question and answer based search result adjustment scheme.
  • the search word from the user is a question and answer type search word
  • the question and answer class result corresponding to the search word is matched in the question and answer knowledge base, and then the search result is performed by using the question and answer class result.
  • the question and answer knowledge base can provide specialized knowledge and information related to the question and answer class, the information provided is large, rich and targeted.
  • the Q&A knowledge base matches the question and answer class results corresponding to the search term, and uses the question and answer class result to search.
  • the result is adjusted to provide the adjusted search result to the user, so that the search result provided to the user is more accurate and comprehensive, so that the search result in the related technology is sorted by the relevance of the search word, and the result is disorderly and unintuitive, and the user selects the result. Difficult, less efficient problems.
  • the present invention can automatically analyze the search term if the search term is a question When answering a class search term, it directly enters the professional Q&A knowledge base to search for the purpose of better understanding the user's query requirements and enhance the user's search experience.
  • the present invention provides the search result including the result of the question and answer class to the user, so that the user can obtain the answer or related knowledge of the question most intuitively and conveniently, and simplifies the work of the user to filter the messy search result by himself.
  • step S102 whether the search word is a question-and-answer type search word, the present invention provides a preferred solution, in which the search word is matched in the question vocabulary, and if there is a question word matching the search word, it is determined The search term is a question and answer class search term. Conversely, if there is no interrogative word matching the search term, it is determined that the search term is not a question-and-answer search term.
  • the question lexicon here stores the question words obtained through data mining or manual configuration, such as "what”, “what”, “how”, “why”, “how”, “practice”, “method”, “The way” and so on.
  • the search result mentioned in step S106 may be a natural search result searched according to the search term in the webpage knowledge base, that is, the natural search result obtained by searching in the webpage knowledge base according to the search term is adjusted by using the question and answer class result.
  • the natural search result obtained through the webpage knowledge base search is a webpage related to the search term, and is sorted by the relevance of the search words, and the user needs further screening to obtain the information of the demand, which is inefficient.
  • the invention provides an adjustment of the natural search result by using the question and answer result, which can obtain the answer or related knowledge of the question most intuitively and conveniently, and simplifies the work of the user to filter the messy search result by himself.
  • the search result may be determined according to the similarity, that is, the search item whose natural search result is similar to the search word is less than or equal to the preset threshold, and the similarity with the search word is less than or A search entry equal to the preset threshold is replaced with a question-and-answer class result.
  • the search term entered by the user is "the practice of tomato eggs”
  • the natural search results obtained by searching in the web knowledge base are "the practice of tomato eggs”, “the practice of scrambled eggs by tomatoes”, “the price of scrambled eggs of tomatoes”, etc.
  • FIG. 2 is a schematic diagram showing search results of the related art in the related art as “the practice of tomato eggs”.
  • FIG. 3 is a schematic diagram showing search results of a search term "the practice of tomato eggs” provided by the technical solution of the present invention.
  • the plurality of question and answer class entries from the same Q&A knowledge base may be combined into one integrated question and answer result before the search result is adjusted by using the question and answer class result in step S106.
  • the search results are adjusted using the integrated question and answer results, and the adjusted search results are provided to the user.
  • resource address data of each question and answer class entry may be extracted, thereby combining multiple question and answer class entries from the same question and answer knowledge base into one integrated question and answer result based on the resource address data.
  • one or more of the plurality of question and answer class entries in the same question and answer knowledge base may be selected as an explicit question and answer result, and the remaining question and answer class entries may be converted into a search link.
  • the elements are integrated with the explicit question and answer results to generate an integrated question and answer result, wherein the link element has link information for initiating a search request based on the search term to the question and answer knowledge base.
  • the search term is “the practice of tomato eggs”.
  • the result of the question and answer, the remaining question and answer class items are converted into a search link element "more related questions” and the explicit question and answer results are integrated to generate an integrated question and answer result, wherein the link element has a search to the question and answer knowledge base.
  • the word initiates the link information for the search request.
  • select a question and answer class entry as the main question in the explicit question and answer results.
  • Answer the result "How to answer the tomato egg _360 question and answer”
  • the identity is integrated with the main question and answer results to generate explicit question and answer results.
  • the jump to the 360 Q&A knowledge base initiates a search request based on the search term, as shown in FIG.
  • the embodiment of the present invention further provides a question and answer based search result adjusting apparatus to implement the above question and answer based search result adjusting method.
  • FIG. 6 is a block diagram showing the structure of a question-and-answer based search result adjusting apparatus according to an embodiment of the present invention.
  • the apparatus may include at least a data input unit 610, an analyzer 620, a matcher 630, a adjuster 640, and a data outputter 650.
  • a data inputter 610 adapted to receive a search term from a user
  • the analyzer 620 is coupled to the data input unit 610 and is adapted to analyze whether the search term is a question-and-answer search term;
  • the matcher 630 is coupled to the analyzer 620, and is adapted to match the Q&A class corresponding to the search term in the Q&A knowledge base when the search term is a Q&A search term;
  • the adjuster 640 is coupled to the matcher 630, and is adapted to adjust the search result by using the result of the question and answer class;
  • a data exporter 650 coupled to the adjuster 640, is adapted to provide the adjusted search results to the user.
  • the adjuster 640 is further adapted to: adjust the natural search result searched in the webpage knowledge base according to the search term by using the question and answer class result.
  • the adjuster 640 is further configured to: determine a search item in the natural search result whose similarity with the search term is less than or equal to a preset threshold; and search for a similarity to the search term that is less than or equal to a preset threshold The entry is replaced with the result of the question and answer class.
  • FIG. 7 is a block diagram showing the structure of a question-and-answer based search result adjusting apparatus according to another embodiment of the present invention.
  • the adjuster 640 may further include: an integrator 710 coupled with the matcher 630 before using the question and answer class result to adjust the search result, suitable for Multiple Q&A entries from the same Q&A knowledge base are combined into one integrated Q&A result.
  • the integrator 710 is further adapted to: select one or more of the plurality of question and answer class entries from the same question and answer knowledge base as an explicit question and answer result; convert the remaining question and answer class entries into A search link element is integrated with the explicit question and answer result to generate an integrated question and answer result, wherein the link element has link information for initiating a search request according to the search word to the question and answer knowledge base.
  • the analyzer 620 is further adapted to: match the search term in the question lexicon, wherein the question vocabulary stores the question word obtained by the data mining method or the manual configuration manner; if there is a matching search word The question word determines that the search term is a question-and-answer search term.
  • the embodiment of the present invention can achieve the following beneficial effects:
  • the invention provides a question and answer based search result adjustment scheme.
  • the search word from the user is a question and answer type search word
  • the question and answer class result corresponding to the search word is matched in the question and answer knowledge base, and then the search result is performed by using the question and answer class result. Adjust to provide the adjusted search results to the user.
  • the Q&A knowledge base can provide specialized Q&A-related knowledge information, the information provided is large, rich and targeted, and the Q&A knowledge base matching search in the present invention
  • the result of the question and answer class corresponding to the word, and the result of the question and answer class is used to adjust the search result, and the adjusted search result is provided to the user, so that the search result provided to the user is more accurate and comprehensive, thereby solving the search result in the related art to search for the word
  • the relevance of the sorting, the results are messy and unintuitive, the user is difficult to screen, and the efficiency is low.
  • the present invention can automatically analyze the search term.
  • the search term is a question-and-answer search term
  • the user directly enters the professional Q&A knowledge base to perform a search to achieve a better understanding of the user's query requirements. Purpose to improve the user's search experience.
  • the present invention provides the search result including the result of the question and answer class to the user, so that the user can obtain the answer or related knowledge of the question most intuitively and conveniently, and simplifies the work of the user to filter the messy search result by himself.
  • an embodiment of the present invention further provides a search result integration method.
  • FIG. 8 shows a flow chart of a search result integration method in accordance with one embodiment of the present invention. As shown in FIG. 8, the method includes at least the following steps S802 to S806.
  • Step S802 receiving a question and answer type search term from a user, and searching for a plurality of search items matching the question and answer type search term.
  • Step S804 extracting a plurality of question and answer class entries in the plurality of search items.
  • Step S806 combining multiple question and answer class entries from the same Q&A knowledge base into one integrated question and answer result.
  • the search result integration method receives a question and answer type search word from a user, searches for a plurality of search items matching the question and answer type search word, and then extracts multiple question and answer class items in the plurality of search items, which will be from the same question and answer.
  • Multiple Q&A entries in the Knowledge Base are combined into one integrated Q&A result. Therefore, the present invention can understand the user's search intention, that is, can identify the question and answer type search words input by the user, and then extract the question and answer class items from the search items matching the question and answer type search words, thereby realizing the purpose of active screening and improving the user search experience. .
  • the present invention can flexibly and intelligently combine multiple question and answer class entries from the same Q&A knowledge base according to the source of the data resource, and realize classification according to different Q&A knowledge bases, and can be provided by the class when provided to the user. Further, in this class, the user is provided with more and more detailed question and answer class entries, and the amount of information provided to the user is large.
  • multiple question and answer class entries before the merge are displayed on the search result page in the form of one or more integrated question and answer results, thereby saving the location resources of the search result page. Further, the saved location is able to display other search results, so additional information can be provided to the user in the same search results page for additional help.
  • the multiple question and answer class entries in the plurality of search items extracted in the above step S804 can be implemented in various manners, such as searching through the back-end question and answer knowledge base or searching for the keyword of the item. These two methods are now described in detail.
  • the background Q & A knowledge base search method That is, multiple Q&A entries in multiple search entries can be extracted through the back-end Q&A knowledge base, wherein the Q&A knowledge base establishes a Q&A data entry through data mining. That is to say, the Q&A data entry can be established in the back-end Q&A knowledge base through data mining in advance for the query to use.
  • Method 2 the way to search for entry keywords. That is, multiple question and answer class entries in the plurality of search entries may be extracted through the presence or absence of an identifier indicating that the search entry is a question and answer class entry among the plurality of search entries. For example, by analyzing whether one or more corresponding Uniform Resource Locators (URLs) in the search entry include a Wenda (question and answer) field, if the Wenda field is included, it is determined that the search entry is a question and answer class entry. For another example, by analyzing whether the question (answer) field in the search entry contains a question and answer, if it is included, it is determined that the search item is a question and answer class entry. Conversely, if it is not included, it is determined that the search entry is not a question-and-answer class entry.
  • URLs Uniform Resource Locators
  • the present invention can extract multiple question and answer class entries in multiple search entries.
  • the resource address data at this time, multiple question and answer class entries from the same Q&A knowledge base can be combined into one integrated question and answer result based on the resource address data.
  • multiple Q&A class entries are from the Q&A knowledge base A, the Q&A knowledge base B, and the Q&A knowledge base C.
  • multiple Q&A entries from the Q&A knowledge base A can be combined into one integrated question and answer result based on the resource address data.
  • Multiple Q&A entries from Q&A B are combined into one integrated Q&A result
  • multiple Q&A entries from Q&A C are combined into one integrated Q&A result.
  • step S806 can be implemented by using one or more of the multiple question and answer class entries from the same question and answer knowledge base as an explicit question and answer result, and the rest.
  • the question and answer class entry is converted into a search link element and merged with the explicit question and answer result to generate an integrated question and answer result, wherein the link element has link information for initiating a search request according to the search word to the question and answer knowledge base.
  • one or more of the plurality of question and answer class entries of the same question and answer knowledge base may be selected as an explicit question and answer result.
  • the Q&A class entries can be sorted according to the user's click volume.
  • the question and answer class entry with the highest user click volume can be used as the displayed question and answer result.
  • one or more of the questions may be selected as an explicit question and answer result, and the user evaluation information may be quantified and converted into a value recognizable by the program. According to the preset selection strategy, it is selected as an explicit question and answer result.
  • a question and answer class entry is selected as the main question and answer result in the explicit question and answer result, and the remaining question and answer class entries are integrated with the main question and answer result as the subordinate question and answer result to generate an explicit question and answer. result. Further, part or all of the answer information of the main question and answer result may be displayed, and the subordinate question and answer result is displayed as a text link.
  • the search result page is drawn by using the integrated question and answer result and the unintegrated search items of the plurality of search items, and is provided to the user, and the search result is obtained in the form of one or more integrated question and answer results after the combination.
  • the page displays multiple Q&A class entries before the merge, saving the location resources of the search results page.
  • the user-entered question-and-answer type search term is “the practice of mashed potatoes”, and the plurality of search items that are searched for matching the search-question-type search term have multiple question-and-answer items, as shown in FIG. Select four of them as explicit question and answer results, and convert the remaining question and answer class entries into a search link element "more related questions" and integrate the explicit question and answer results to generate an integrated question and answer result, where the link elements A link message with a search request initiated by the question and answer knowledge base based on the search term.
  • the user-entered question-and-answer search term is “nutrition value of the egg”, and the search results are matched with the question-and-answer search term from the question-answer knowledge base A and the question-and-answer knowledge base.
  • B multiple question and answer class entries at this time, the multiple question and answer class entries from the Q&A knowledge base A are merged into one integrated question and answer result a, and the multiple question and answer class entries from the Q&A knowledge base B are combined into one integrated question and answer result b,
  • the search result page is drawn by using the integrated question and answer results a and b and the unintegrated search items of the plurality of search items, and is provided to the user, as shown in FIG.
  • an embodiment of the present invention further provides a search result integration apparatus to implement the foregoing search result integration method.
  • FIG. 12 is a block diagram showing the structure of a search result integration apparatus according to an embodiment of the present invention.
  • the apparatus may include at least: a data inputter 1210, a searcher 1220, an extractor 1230, and an integrator 1240.
  • a data inputter 1210 adapted to receive a question and answer search term from a user
  • the searcher 1220 coupled to the data inputter 1210, is adapted to search for a plurality of search entries that match the search term search term;
  • An extractor 1230 coupled to the searcher 1220, is adapted to extract a plurality of question and answer class entries of the plurality of search terms;
  • the integrator 1240 coupled to the extractor 1230, is adapted to combine multiple question and answer class entries from the same question and answer knowledge base into one integrated question and answer result.
  • the extractor 1230 is further configured to: extract resource address data of multiple question and answer class entries of the plurality of search entries;
  • the integrator 1240 is further adapted to merge the plurality of question and answer class entries from the same Q&A knowledge base into an integrated question and answer result based on the resource address data.
  • FIG. 13 is a block diagram showing the structure of a search result integration apparatus according to another embodiment of the present invention.
  • the integrator 1240 combines a plurality of question and answer class entries from the same Q&A knowledge base into an integrated question and answer result, and further includes: a data exporter 1310 coupled with the integrator 1240, adapted to utilize the integrated question and answer results and Search terms that are not integrated in the search terms are drawn and provided to the user.
  • the integrator 1240 is further adapted to: select one or more of the plurality of question and answer class entries from the same question and answer knowledge base as an explicit question and answer result; convert the remaining question and answer class entries into one
  • the search link element is merged with the explicit question and answer result to generate an integrated question and answer result, wherein the link element has link information for initiating a search request according to the search word to the question and answer knowledge base.
  • the integrator 1240 is further adapted to: select one or more of the user clicks of the plurality of question and answer class entries of the same question and answer knowledge base as an explicit question and answer result.
  • the integrator 1240 is further adapted to: select a question and answer class entry as the main question and answer result in the explicit question and answer result; and merge the remaining question and answer class entries with the primary question and answer result as the primary question and answer result to generate the result. Explicit question and answer results.
  • the integrator 1240 is further adapted to: display part or all of the answer information of the main question and answer result, and display the subordinate question and answer result in a text link manner.
  • the extractor 1230 is further configured to: extract, by using a back-end question and answer knowledge base, a plurality of question-and-answer class entries in the plurality of search items, wherein the question-and-answer knowledge base establishes a question-and-answer class data item through data mining; or Whether there is an identifier in the search term indicating that the search entry is a question-and-answer class entry, and multiple question-answer class entries in the plurality of search entries are extracted.
  • the embodiment of the present invention can achieve the following beneficial effects:
  • the search result integration method receives a question and answer type search word from a user, searches for a plurality of search items matching the question and answer type search word, and then extracts multiple question and answer class items in the plurality of search items, which will be from the same Multiple Q&A entries in a Q&A knowledge pool are combined into one integrated Q&A result. Therefore, the present invention can understand the user's search intention, that is, can identify the question and answer type search words input by the user, and then extract the question and answer class items from the search items matching the question and answer type search words, thereby realizing the purpose of active screening and improving the user search experience. .
  • the present invention can flexibly and intelligently combine multiple question and answer class entries from the same Q&A knowledge base according to the source of the data resource, and realize classification according to different Q&A knowledge bases, and can be provided by the class when provided to the user. Further, in this class, the user is provided with more and more detailed question and answer class entries, and the amount of information provided to the user is large.
  • multiple question and answer class entries before the merge are displayed on the search result page in the form of one or more integrated question and answer results, thereby saving the location resources of the search result page. Further, the saved location is able to display other search results, so additional information can be provided to the user in the same search results page for additional help.
  • modules in the devices of the embodiments can be adaptively changed and placed in one or more devices different from the embodiment.
  • the modules or units or components of the embodiments may be combined into one module or unit or component, and further they may be divided into a plurality of sub-modules or sub-units or sub-components.
  • any combination of the features disclosed in the specification, including the accompanying claims, the abstract and the drawings, and any methods so disclosed, or All processes or units of the device are combined.
  • Each feature disclosed in this specification (including the accompanying claims, the abstract and the drawings) may be replaced by alternative features that provide the same, equivalent or similar purpose.
  • the various component embodiments of the present invention may be implemented in hardware, or in a software module running on one or more processors, or in a combination thereof.
  • a microprocessor or digital signal processor can be used in practice to implement some or all of the Q&A-based search result adjustment apparatus and search result integration apparatus in accordance with embodiments of the present invention.
  • the invention can also be implemented as a device or device program (e.g., a computer program and a computer program product) for performing some or all of the methods described herein.
  • a program implementing the invention may be stored on a computer readable medium or may be in the form of one or more signals. Such signals may be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.
  • FIG. 14 illustrates a computing device that can implement a question and answer based search result adjustment method and a search result integration method.
  • the computing device conventionally includes a processor 1410 and a computer program product or computer readable medium in the form of a memory 1420.
  • the memory 1420 may be an electronic memory such as a flash memory, an EEPROM (Electrically Erasable Programmable Read Only Memory), an EPROM, a hard disk, or a ROM.
  • Memory 1420 has a memory space 1430 for program code 1431 for performing any of the method steps described above.
  • storage space 1430 for program code may include various program code 1431 for implementing various steps in the above methods, respectively.
  • the program code can be read from or written to one or more computer program products.
  • Such computer program products include program code carriers such as hard disks, compact disks (CDs), memory cards or floppy disks.
  • Such a computer program product is typically a portable or fixed storage unit as described with reference to FIG.
  • the storage unit may have storage segments, storage spaces, and the like that are similarly arranged to memory 1420 in the computing device of FIG.
  • the program code can be compressed, for example, in an appropriate form.
  • the storage unit includes computer readable code 1431', ie, code that can be read by, for example, a processor such as 1410, which when executed by the computing device causes the computing device to perform each of the methods described above step.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

本发明公开了一种基于问答的搜索结果调整方法和装置，以及一种搜索结果整合方法和装置。该基于问答的搜索结果调整方法包括：接收来自用户的搜索词，分析所述搜索词是否为问答类搜索词；当所述搜索词为问答类搜索词时，在问答知识库中匹配所述搜索词对应的问答类结果；利用所述问答类结果对搜索结果进行调整，将调整后的搜索结果提供给所述用户。本发明中在问答知识库匹配搜索词对应的问答类结果，并利用问答类结果对搜索结果进行调整，将调整后的搜索结果提供给用户，使得提供给用户的搜索结果更加准确、丰富及全面。

Description

基于问答的搜索结果调整方法和装置 技术领域
本发明涉及信息搜索领域，特别是一种基于问答的搜索结果调整方法和装置，以及一种搜索结果整合方法和装置。
背景技术
随着信息技术的飞速发展，当今社会进入了信息***时代，人们越来越多地借助网络来寻找自己需要的信息，因此，检索成为人们工作、生活不可或缺的一部分。人们通常使用搜索引擎来进行检索，搜索引擎是指根据一定的策略、运用特定的计算机程序从互联网上搜集信息，在对信息进行组织和处理后，为用户提供检索服务，将与用户检索相关的信息展示给用户的***。
相关技术中，基于用户的提问(即用户输入问答类搜索词)，现有搜索引擎给出的搜索结果以搜索词的相关性进行排序，结果杂乱无章不直观，用户筛选起来困难，效率较低，也非常不便捷。
发明内容
鉴于上述问题，提出了本发明以便提供一种克服上述问题或者至少部分地解决上述问题的基于问答的搜索结果调整方法和装置，以及搜索结果整合方法和装置。
根据本发明的一方面，提供了一种基于问答的搜索结果调整方法，包括：接收来自用户的搜索词，分析所述搜索词是否为问答类搜索词；当所述搜索词为问答类搜索词时，在问答知识库中匹配所述搜索词对应的问答类结果；利用所述问答类结果对搜索结果进行调整，将调整后的搜索结果提供给所述用户。
根据本发明的另一方面，还提供了一种基于问答的搜索结果调整装置，包括：数据输入器，适于接收来自用户的搜索词；分析器，适于分析所述搜索词是否为问答类搜索词；匹配器，适于当所述搜索词为问答类搜索词时，在问答知识库中匹配所述搜索词对应的问答类结果；调整器，适于利用所述问答类结果对搜索结果进行调整；数据输出器，适于将调整后的搜索结果提供给所述用户。
本发明的有益效果为：
本发明提供了一种基于问答的搜索结果调整方案，当来自用户的搜索词为问答类搜索词时，在问答知识库中匹配搜索词对应的问答类结果，进而利用问答类结果对搜索结果进行调整，将调整后的搜索结果提供给用户。由于问答知识库能够提供专门的问答类相关知识信息，其提供的信息量大、丰富且有针对性，本发明中在问答知识库匹配搜索词对应的问答类结果，并利用问答类结果对搜索结果进行调整，将调整后的搜索结果提供给用户，使得提供给用户的搜索结果更加准确、全面，从而解决相关技术中搜索结果以搜索词的相关性进行排序，结果杂乱无章不直观，用户筛选起来困难，效率较低的问题。并且，当接收到来自用户的搜索词时，本发明能够自动分析搜索词，若搜索词为问答类搜索词时，则直接进入专业的问答知识库进行查找，实现更好地理解用户查询需求的目的，提升用户的搜索体验。此外，本发明将包含问答类结果的搜索结果提供给用户， 使得用户可以最直观、便捷地获得问题的答案或相关知识，简化了用户自行筛选杂乱的搜索结果的工作。
根据本发明的另一方面，还提供了一种搜索结果整合方法，包括：接收来自用户的问答类搜索词，搜索得到与所述问答类搜索词匹配的多条搜索条目；提取所述多条搜索条目中的多条问答类条目；将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。
根据本发明的另一方面，还提供了一种搜索结果整合装置，包括：数据输入器，适于接收来自用户的问答类搜索词；搜索器，适于搜索得到与所述问答类搜索词匹配的多条搜索条目；提取器，适于提取所述多条搜索条目中的多条问答类条目；整合器，适于将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。
本发明的有益效果为：
本发明提供的搜索结果整合方法，接收来自用户的问答类搜索词，搜索得到与问答类搜索词匹配的多条搜索条目，随后提取多条搜索条目中的多条问答类条目，将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。由此，本发明能够理解用户的搜索意图，即能够识别出用户输入的问答类搜索词，进而从匹配问答类搜索词的搜索条目中提取问答类条目，实现主动筛选的目的，提升用户搜索体验。并且，本发明能够依据数据资源的来源灵活地、智能地将来自同一问答知识库的多条问答类条目合并，实现按照不同的问答知识库进行分类，提供给用户的时候能够以类来提供，进一步在该类中为用户提供更多更细的问答类条目，提供给用户的信息量大。合并后以一条或多条整合问答结果的形式在搜索结果页面展示合并前的多条问答类条目，从而节省了搜索结果页面的位置资源。进一步地，节省的位置能够显示其它搜索结果，因此能够在同一搜索结果页面中为用户提供更多的信息，实现额外帮助。
根据本发明的又一方面，提供了一种计算机程序，其包括计算机可读代码，当所述计算机可读代码在计算设备上运行时，导致所述计算设备执行根据上文所述的基于问答的搜索结果调整方法以及搜索结果整合方法。
根据本发明的再一方面，提供了一种计算机可读介质，其中存储了上述的计算机程序。
上述说明仅是本发明技术方案的概述，为了能够更清楚了解本发明的技术手段，而可依照说明书的内容予以实施，并且为了让本发明的上述和其它目的、特征和优点能够更明显易懂，以下特举本发明的具体实施方式。
附图说明
通过阅读下文优选实施方式的详细描述，各种其他的优点和益处对于本领域普通技术人员将变得清楚明了。附图仅用于示出优选实施方式的目的，而并不认为是对本发明的限制。而且在整个附图中，用相同的参考符号表示相同的部件。在附图中：
图1示出了根据本发明一个实施例的基于问答的搜索结果调整方法的流程图；
图2示出了相关技术中搜索词为“西红柿鸡蛋的做法”的搜索结果示意图；
图3示出了采用本发明的技术方案提供的搜索词为“西红柿鸡蛋的做法”的一种搜索结果示意图；
图4示出了采用本发明的技术方案提供的搜索词为“西红柿鸡蛋的做法”的另一种搜索结果示意图；
图5示出了点击链接元素“更多相关问题”时跳转至问答知识库依据搜索词发起搜索请求的结果示意图；
图6示出了根据本发明一个实施例的基于问答的搜索结果调整装置结构示意图；
图7示出了根据本发明另一个实施例的基于问答的搜索结果调整装置的结构示意图；
图8示出了根据本发明一个实施例的搜索结果整合方法的流程图；
图9示出了采用本发明的技术方案提供的问答类搜索词为“土豆泥的做法”的搜索结果示意图；
图10示出了点击链接元素“更多相关问题”时跳转至问答知识库依据搜索词发起搜索请求的结果示意图；
图11示出了采用本发明的技术方案提供的问答类搜索词为“鸡蛋的营养价值”的搜索结果示意图；
图12示出了根据本发明一个实施例的搜索结果整合装置结构示意图；
图13示出了根据本发明另一个实施例的搜索结果整合装置结构示意图；
图14示意性地示出了用于执行根据本发明的基于问答的搜索结果调整方法以及搜索结果整合方法的计算设备的框图；以及
图15示意性地示出了用于保持或者携带实现根据本发明的基于问答的搜索结果调整方法以及搜索结果整合方法的程序代码的存储单元。
具体实施方式
下面结合附图和具体的实施方式对本发明作进一步的描述。
为解决上述技术问题，本发明实施例提供了一种基于问答的搜索结果调整方法。图1示出了根据本发明一个实施例的基于问答的搜索结果调整方法的流程图。如图1所示，该方法至少包括以下步骤S102至步骤S106。
步骤S102、接收来自用户的搜索词，分析搜索词是否为问答类搜索词。
步骤S104、当搜索词为问答类搜索词时，在问答知识库中匹配搜索词对应的问答类结果。
步骤S106、利用问答类结果对搜索结果进行调整，将调整后的搜索结果提供给用户。
本发明提供了一种基于问答的搜索结果调整方案，当来自用户的搜索词为问答类搜索词时，在问答知识库中匹配搜索词对应的问答类结果，进而利用问答类结果对搜索结果进行调整，将调整后的搜索结果提供给用户。由于问答知识库能够提供专门的问答类相关知识信息，其提供的信息量大、丰富且有针对性，本发明中在问答知识库匹配搜索词对应的问答类结果，并利用问答类结果对搜索结果进行调整，将调整后的搜索结果提供给用户，使得提供给用户的搜索结果更加准确、全面，从而解决相关技术中搜索结果以搜索词的相关性进行排序，结果杂乱无章不直观，用户筛选起来困难，效率较低的问题。并且，当接收到来自用户的搜索词时，本发明能够自动分析搜索词，若搜索词为问 答类搜索词时，则直接进入专业的问答知识库进行查找，实现更好地理解用户查询需求的目的，提升用户的搜索体验。此外，本发明将包含问答类结果的搜索结果提供给用户，使得用户可以最直观、便捷地获得问题的答案或相关知识，简化了用户自行筛选杂乱的搜索结果的工作。
上文步骤S102中分析搜索词是否为问答类搜索词，本发明提供了一种优选的方案，在该方案中，在疑问词库中匹配搜索词，若存在匹配搜索词的疑问词，则确定搜索词为问答类搜索词。反之，若不存在匹配搜索词的疑问词，则确定搜索词不为问答类搜索词。这里的疑问词库中存储了通过数据挖掘方式或人工配置方式得到的疑问词，如“哪些”、“什么”、“如何”、“为何”、“怎么”、“做法”、“方法”、“方式”等等。
进一步地，步骤S106提及的搜索结果可以是根据搜索词在网页知识库中搜索得到的自然搜索结果，即利用问答类结果对根据搜索词在网页知识库中搜索得到的自然搜索结果进行调整。一般而言，通过网页知识库搜索得到的自然搜索结果是与搜索词相关的网页，并以搜索词的相关性进行排序，用户还需要进一步筛选来获取需求的信息，效率较低。通过本发明提供利用问答类结果对的自然搜索结果进行调整，可以最直观、便捷地获得问题的答案或相关知识，简化了用户自行筛选杂乱的搜索结果的工作。在利用问答类结果对自然搜索结果进行调整时，可以根据相似度进行，即判断自然搜索结果中与搜索词的相似度小于或等于预设阈值的搜索条目，将与搜索词的相似度小于或等于预设阈值的搜索条目替换为问答类结果。例如，用户输入的搜索词为“西红柿鸡蛋的做法”，在网页知识库中搜索得到的自然搜索结果为“西红柿鸡蛋的做法”、“西红柿炒鸡蛋的做法”、“西红柿炒鸡蛋的价钱”等等，通过分析确定搜索条目“西红柿炒鸡蛋的价钱”与搜索词“西红柿鸡蛋的做法”的相似度小于预设阈值，此时可以将该搜索条目替换为问答类结果。如图2所示为相关技术中搜索词为“西红柿鸡蛋的做法”的搜索结果示意图。图3示出了采用本发明的技术方案提供的搜索词为“西红柿鸡蛋的做法”的搜索结果示意图。通过对比图2和图3，得到两者的区别地方在虚线标记的部分，采用图3中虚线标记的搜索条目替换掉图2中虚线标记的搜索条目，且图3中虚线标记的搜索条目直接将问题的答案或相关知识作为搜索结果提供给用户，简化了用户筛选的工作。
若问答类结果中问答类条目的数量为多条，则步骤S106利用问答类结果对搜索结果进行调整之前，还可以将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。也就是说，利用整合问答结果对搜索结果进行调整，将调整后的搜索结果提供给用户。具体地，可以提取每条问答类条目的资源地址数据，从而基于资源地址数据将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。在本发明的优选方案中，对于来自同一问答知识库中的多条问答类条目，可以选取其中的一条或多条作为显式的问答结果，将其余的问答类条目转化为一搜索用的链接元素与显式的问答结果整合，生成一条整合问答结果，其中，链接元素带有向该问答知识库依据搜索词发起搜索请求的链接信息。仍以搜索词为“西红柿鸡蛋的做法”为例，在问答知识库中匹配该搜索词对应的问答类结果中有多条问答类条目，如图4所示，选取其中的三条作为显式的问答结果，将其余的问答类条目转化为一搜索用的链接元素“更多相关问题”与显式的问答结果整合，生成一条整合问答结果，其中，链接元素带有向该问答知识库依据搜索词发起搜索请求的链接信息。如图4所示，在显式的问答结果中选取一条问答类条目作为主问 答结果“西红柿鸡蛋的做法_360问答”，展示该主问答结果的部分或全部的答案信息，将其余的问答类条目“西红柿鸡蛋汤的做法”、“西红柿鸡蛋面的做法”以从属问答结果的身份与主问答结果整合，生成显式的问答结果。进一步地，当用户点击链接元素“更多相关问题”时，则跳转至360问答知识库依据搜索词发起搜索请求，如图5所示。
基于同一发明构思，本发明实施例还提供了一种基于问答的搜索结果调整装置，以实现上述基于问答的搜索结果调整方法。
图6示出了根据本发明一个实施例的基于问答的搜索结果调整装置的结构示意图。参见图6，该装置至少可以包括：数据输入器610、分析器620、匹配器630、调整器640以及数据输出器650。
现介绍本发明实施例的基于问答的搜索结果调整装置的各组成或器件的功能以及各部分间的连接关系：
数据输入器610，适于接收来自用户的搜索词；
分析器620，与数据输入器610相耦合，适于分析搜索词是否为问答类搜索词；
匹配器630，与分析器620相耦合，适于当搜索词为问答类搜索词时，在问答知识库中匹配搜索词对应的问答类结果；
调整器640，与匹配器630相耦合，适于利用问答类结果对搜索结果进行调整；
数据输出器650，与调整器640相耦合，适于将调整后的搜索结果提供给用户。
在一个实施例中，上述调整器640还适于：利用问答类结果对根据搜索词在网页知识库中搜索得到的自然搜索结果进行调整。
在一个实施例中，上述调整器640还适于：判断自然搜索结果中与搜索词的相似度小于或等于预设阈值的搜索条目；将与搜索词的相似度小于或等于预设阈值的搜索条目替换为问答类结果。
在一个实施例中，图7示出了根据本发明另一个实施例的基于问答的搜索结果调整装置的结构示意图。参见图7，若问答类结果中问答类条目的数量为多条，调整器640利用问答类结果对搜索结果进行调整之前，还可以包括：整合器710，与匹配器630相耦合，适于将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。
在一个实施例中，上述整合器710还适于：对于来自同一问答知识库中的多条问答类条目，选取其中的一条或多条作为显式的问答结果；将其余的问答类条目转化为一搜索用的链接元素与显式的问答结果整合，生成一条整合问答结果，其中，链接元素带有向该问答知识库依据搜索词发起搜索请求的链接信息。
在一个实施例中，上述分析器620还适于：在疑问词库中匹配搜索词，其中，疑问词库中存储了通过数据挖掘方式或人工配置方式得到的疑问词；若存在匹配搜索词的疑问词，则确定搜索词为问答类搜索词。
根据上述任意一个优选实施例或多个优选实施例的组合，本发明实施例能够达到如下有益效果：
本发明提供了一种基于问答的搜索结果调整方案，当来自用户的搜索词为问答类搜索词时，在问答知识库中匹配搜索词对应的问答类结果，进而利用问答类结果对搜索结果进行调整，将调整后的搜索结果提供给用户。由于问答知识库能够提供专门的问答类相关知识信息，其提供的信息量大、丰富且有针对性，本发明中在问答知识库匹配搜索 词对应的问答类结果，并利用问答类结果对搜索结果进行调整，将调整后的搜索结果提供给用户，使得提供给用户的搜索结果更加准确、全面，从而解决相关技术中搜索结果以搜索词的相关性进行排序，结果杂乱无章不直观，用户筛选起来困难，效率较低的问题。并且，当接收到来自用户的搜索词时，本发明能够自动分析搜索词，若搜索词为问答类搜索词时，则直接进入专业的问答知识库进行查找，实现更好地理解用户查询需求的目的，提升用户的搜索体验。此外，本发明将包含问答类结果的搜索结果提供给用户，使得用户可以最直观、便捷地获得问题的答案或相关知识，简化了用户自行筛选杂乱的搜索结果的工作。
根据本发明的另一实施例，本发明实施例还提供了一种搜索结果整合方法。图8示出了根据本发明一个实施例的搜索结果整合方法的流程图。如图8所示，该方法至少包括以下步骤S802至步骤S806。
步骤S802、接收来自用户的问答类搜索词，搜索得到与问答类搜索词匹配的多条搜索条目。
步骤S804、提取多条搜索条目中的多条问答类条目。
步骤S806、将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。
本发明提供的搜索结果整合方法，接收来自用户的问答类搜索词，搜索得到与问答类搜索词匹配的多条搜索条目，随后提取多条搜索条目中的多条问答类条目，将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。由此，本发明能够理解用户的搜索意图，即能够识别出用户输入的问答类搜索词，进而从匹配问答类搜索词的搜索条目中提取问答类条目，实现主动筛选的目的，提升用户搜索体验。并且，本发明能够依据数据资源的来源灵活地、智能地将来自同一问答知识库的多条问答类条目合并，实现按照不同的问答知识库进行分类，提供给用户的时候能够以类来提供，进一步在该类中为用户提供更多更细的问答类条目，提供给用户的信息量大。合并后以一条或多条整合问答结果的形式在搜索结果页面展示合并前的多条问答类条目，从而节省了搜索结果页面的位置资源。进一步地，节省的位置能够显示其它搜索结果，因此能够在同一搜索结果页面中为用户提供更多的信息，实现额外帮助。
上文步骤S804中提取多条搜索条目中的多条问答类条目可以采用多种方式来实现，如通过后台问答知识库搜索或者搜索条目关键词等方式。现对这两种方式进行详细介绍。
方式一，后台问答知识库搜索的方式。即可以通过后台问答知识库提取多条搜索条目中的多条问答类条目，其中，问答知识库通过数据挖掘建立问答类数据条目。也就是说，可以预先通过数据挖掘在后台问答知识库中建立问答类数据条目，以供查询使用。
方式二，搜索条目关键词的方式。即可以通过多条搜索条目中是否存在表明该搜索条目为问答类条目的标识，提取多条搜索条目中的多条问答类条目。例如，通过分析搜索条目中的一个或多个对应的统一资源定位符(Uniform Resource Locator，URL)中是否包含wenda(问答)字段，如果包含wenda字段，则确定该搜索条目为问答类条目。又例如，通过分析搜索条目中的title(标题)字段中是否包含问答，如果包含，则确定该搜索条目为问答类条目。反之，如果不包含，则确定该搜索条目不为问答类条目。
进一步地，在步骤S806之前，本发明可以提取多条搜索条目中的多条问答类条目 的资源地址数据，此时可以基于资源地址数据将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。例如，多条问答类条目来自问答知识库A、问答知识库B以及问答知识库C，此时可以基于资源地址数据将来自问答知识库A的多条问答类条目合并为一条整合问答结果，将来自问答知识库B的多条问答类条目合并为一条整合问答结果，将来自问答知识库C的多条问答类条目合并为一条整合问答结果。
在本发明的优选实施例中，步骤S806可以采用如下技术手段来实现，即对于来自同一问答知识库的多条问答类条目，选取其中的一条或多条作为显式的问答结果，并将其余的问答类条目转化为一搜索用的链接元素与显式的问答结果合并，生成一条整合问答结果，其中，链接元素带有向该问答知识库依据搜索词发起搜索请求的链接信息。进一步地，可以基于同一问答知识库的多条问答类条目的用户点击量，选取其中的一条或多条作为显式的问答结果。可以根据用户点击量对问答类条目进行排序，一般来说，用户点击量越高其对应的问答类条目质量越好，越得到用户的认可。例如可以将用户点击量最高的问答类条目作为显示的问答结果。此外，还可以基于同一问答知识库的多条问答类条目的用户评价信息，选取其中的一条或多条作为显式的问答结果，具体可以将用户评价信息进行量化，转化成程序能够识别的数值，根据预设的选取策略进行选取作为显式的问答结果。在本发明的另一个实施例中，在显式的问答结果中选取一条问答类条目作为主问答结果，将其余的问答类条目以从属问答结果的身份与主问答结果整合，生成显式的问答结果。进一步地，可以展示主问答结果的部分或全部的答案信息，且以文字链接的方式展示从属问答结果。
进而，在步骤S806之后，利用整合问答结果以及多条搜索条目中未被整合的搜索条目绘制搜索结果页，并提供给用户，实现了合并后以一条或多条整合问答结果的形式在搜索结果页面展示合并前的多条问答类条目，从而节省了搜索结果页面的位置资源。
下面通过具体实施例详细介绍本发明提供的搜索结果整合方法。在具体实施例中，用户输入的问答类搜索词为“土豆泥的做法”，在搜索得到的与该问答类搜索词匹配的多条搜索条目中有多条问答类条目，如图9所示，选取其中的四条作为显式的问答结果，将其余的问答类条目转化为一搜索用的链接元素“更多相关问题”与显式的问答结果整合，生成一条整合问答结果，其中，链接元素带有向问答知识库依据搜索词发起搜索请求的链接信息。如图9所示，在显式的问答结果中选取一条问答类条目作为主问答结果“土豆泥的做法_360问答”，展示该主问答结果的部分或全部的答案信息，将其余的问答类条目“土豆泥有几种做法”、“肯德基土豆泥的做法”、“麦当劳土豆泥的做法？”以从属问答结果的身份与主问答结果整合，生成显式的问答结果。进一步地，当用户点击链接元素“更多相关问题”时，则跳转至360问答知识库依据搜索词发起搜索请求，如图10所示。
在另一个具体实施例中，用户输入的问答类搜索词为“鸡蛋的营养价值”，在搜索得到的与该问答类搜索词匹配的多条搜索条目中有来自问答知识库A和问答知识库B的多条问答类条目，此时将来自问答知识库A的多条问答类条目合并为一条整合问答结果a，将来自问答知识库B的多条问答类条目合并为一条整合问答结果b，并利用整合问答结果a和b以及多条搜索条目中未被整合的搜索条目绘制搜索结果页，并提供给用户，如图11所示。
基于同一发明构思，本发明实施例还提供了一种搜索结果整合装置，以实现上述搜索结果整合方法。
图12示出了根据本发明一个实施例的搜索结果整合装置的结构示意图。参见图12，该装置至少可以包括：数据输入器1210、搜索器1220、提取器1230以及整合器1240。
现介绍本发明实施例的搜索结果整合装置的各组成或器件的功能以及各部分间的连接关系：
数据输入器1210，适于接收来自用户的问答类搜索词；
搜索器1220，与数据输入器1210相耦合，适于搜索得到与问答类搜索词匹配的多条搜索条目；
提取器1230，与搜索器1220相耦合，适于提取多条搜索条目中的多条问答类条目；
整合器1240，与提取器1230相耦合，适于将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。
在一个实施例中，上述提取器1230还适于：提取多条搜索条目中的多条问答类条目的资源地址数据；
上述整合器1240还适于：基于资源地址数据将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。
在一个实施例中，图13示出了根据本发明另一个实施例的搜索结果整合装置结构示意图。参见图13，整合器1240将来自同一问答知识库的多条问答类条目合并为一条整合问答结果之后，还包括：数据输出器1310，与整合器1240相耦合，适于利用整合问答结果以及多条搜索条目中未被整合的搜索条目绘制搜索结果页，并提供给用户。
在一个实施例中，上述整合器1240还适于：对于来自同一问答知识库的多条问答类条目，选取其中的一条或多条作为显式的问答结果；将其余的问答类条目转化为一搜索用的链接元素与显式的问答结果合并，生成一条整合问答结果，其中，链接元素带有向该问答知识库依据搜索词发起搜索请求的链接信息。
在一个实施例中，上述整合器1240还适于：基于同一问答知识库的多条问答类条目的用户点击量，选取其中的一条或多条作为显式的问答结果。
在一个实施例中，上述整合器1240还适于：在显式的问答结果中选取一条问答类条目作为主问答结果；将其余的问答类条目以从属问答结果的身份与主问答结果合并，生成显式的问答结果。
在一个实施例中，上述整合器1240还适于：展示主问答结果的部分或全部的答案信息，且以文字链接的方式展示从属问答结果。
在一个实施例中，上述提取器1230还适于：通过后台问答知识库提取多条搜索条目中的多条问答类条目，其中，问答知识库通过数据挖掘建立问答类数据条目；或者，通过多条搜索条目中是否存在表明该搜索条目为问答类条目的标识，提取多条搜索条目中的多条问答类条目。
根据上述任意一个优选实施例或多个优选实施例的组合，本发明实施例能够达到如下有益效果：
本发明提供的搜索结果整合方法，接收来自用户的问答类搜索词，搜索得到与问答类搜索词匹配的多条搜索条目，随后提取多条搜索条目中的多条问答类条目，将来自同 一问答知识库的多条问答类条目合并为一条整合问答结果。由此，本发明能够理解用户的搜索意图，即能够识别出用户输入的问答类搜索词，进而从匹配问答类搜索词的搜索条目中提取问答类条目，实现主动筛选的目的，提升用户搜索体验。并且，本发明能够依据数据资源的来源灵活地、智能地将来自同一问答知识库的多条问答类条目合并，实现按照不同的问答知识库进行分类，提供给用户的时候能够以类来提供，进一步在该类中为用户提供更多更细的问答类条目，提供给用户的信息量大。合并后以一条或多条整合问答结果的形式在搜索结果页面展示合并前的多条问答类条目，从而节省了搜索结果页面的位置资源。进一步地，节省的位置能够显示其它搜索结果，因此能够在同一搜索结果页面中为用户提供更多的信息，实现额外帮助。
在此处所提供的说明书中，说明了大量具体细节。然而，能够理解，本发明的实施例可以在没有这些具体细节的情况下实践。在一些实例中，并未详细示出公知的方法、结构和技术，以便不模糊对本说明书的理解。
类似地，应当理解，为了精简本公开并帮助理解各个发明方面中的一个或多个，在上面对本发明的示例性实施例的描述中，本发明的各个特征有时被一起分组到单个实施例、图、或者对其的描述中。然而，并不应将该公开的方法解释成反映如下意图：即所要求保护的本发明要求比在每个权利要求中所明确记载的特征更多的特征。更确切地说，如下面的权利要求书所反映的那样，发明方面在于少于前面公开的单个实施例的所有特征。因此，遵循具体实施方式的权利要求书由此明确地并入该具体实施方式，其中每个权利要求本身都作为本发明的单独实施例。
本领域那些技术人员可以理解，可以对实施例中的设备中的模块进行自适应性地改变并且把它们设置在与该实施例不同的一个或多个设备中。可以把实施例中的模块或单元或组件组合成一个模块或单元或组件，以及此外可以把它们分成多个子模块或子单元或子组件。除了这样的特征和/或过程或者单元中的至少一些是相互排斥之外，可以采用任何组合对本说明书(包括伴随的权利要求、摘要和附图)中公开的所有特征以及如此公开的任何方法或者设备的所有过程或单元进行组合。除非另外明确陈述，本说明书(包括伴随的权利要求、摘要和附图)中公开的每个特征可以由提供相同、等同或相似目的的替代特征来代替。
此外，本领域的技术人员能够理解，尽管在此所述的一些实施例包括其它实施例中所包括的某些特征而不是其它特征，但是不同实施例的特征的组合意味着处于本发明的范围之内并且形成不同的实施例。例如，在下面的权利要求书中，所要求保护的实施例的任意之一都可以以任意的组合方式来使用。
本发明的各个部件实施例可以以硬件实现，或者以在一个或者多个处理器上运行的软件模块实现，或者以它们的组合实现。本领域的技术人员应当理解，可以在实践中使用微处理器或者数字信号处理器(DSP)来实现根据本发明实施例的基于问答的搜索结果调整装置以及搜索结果整合装置中的一些或者全部部件的一些或者全部功能。本发明还可以实现为用于执行这里所描述的方法的一部分或者全部的设备或者装置程序(例如，计算机程序和计算机程序产品)。这样的实现本发明的程序可以存储在计算机可读介质上，或者可以具有一个或者多个信号的形式。这样的信号可以从因特网网站上下载得到，或者在载体信号上提供，或者以任何其他形式提供。
例如，图14示出了可以实现基于问答的搜索结果调整方法以及搜索结果整合方法的计算设备。该计算设备传统上包括处理器1410和以存储器1420形式的计算机程序产品或者计算机可读介质。存储器1420可以是诸如闪存、EEPROM(电可擦除可编程只读存储器)、EPROM、硬盘或者ROM之类的电子存储器。存储器1420具有用于执行上述方法中的任何方法步骤的程序代码1431的存储空间1430。例如，用于程序代码的存储空间1430可以包括分别用于实现上面的方法中的各种步骤的各个程序代码1431。这些程序代码可以从一个或者多个计算机程序产品中读出或者写入到这一个或者多个计算机程序产品中。这些计算机程序产品包括诸如硬盘，紧致盘(CD)、存储卡或者软盘之类的程序代码载体。这样的计算机程序产品通常为如参考图15所述的便携式或者固定存储单元。该存储单元可以具有与图14的计算设备中的存储器1420类似布置的存储段、存储空间等。程序代码可以例如以适当形式进行压缩。通常，存储单元包括计算机可读代码1431’，即可以由例如诸如1410之类的处理器读取的代码，这些代码当由计算设备运行时，导致该计算设备执行上面所描述的方法中的各个步骤。
本文中所称的“一个实施例”、“实施例”或者“一个或者多个实施例”意味着，结合实施例描述的特定特征、结构或者特性包括在本发明的至少一个实施例中。此外，请注意，这里“在一个实施例中”的词语例子不一定全指同一个实施例。
应该注意的是上述实施例对本发明进行说明而不是对本发明进行限制，并且本领域技术人员在不脱离所附权利要求的范围的情况下可设计出替换实施例。在权利要求中，不应将位于括号之间的任何参考符号构造成对权利要求的限制。单词“包含”不排除存在未列在权利要求中的元件或步骤。位于元件之前的单词“一”或“一个”不排除存在多个这样的元件。本发明可以借助于包括有若干不同元件的硬件以及借助于适当编程的计算机来实现。在列举了若干装置的单元权利要求中，这些装置中的若干个可以是通过同一个硬件项来具体体现。单词第一、第二、以及第三等的使用不表示任何顺序。可将这些单词解释为名称。
此外，还应当注意，本说明书中使用的语言主要是为了可读性和教导的目的而选择的，而不是为了解释或者限定本发明的主题而选择的。因此，在不偏离所附权利要求书的范围和精神的情况下，对于本技术领域的普通技术人员来说许多修改和变更都是显而易见的。对于本发明的范围，对本发明所做的公开是说明性的，而非限制性的，本发明的范围由所附权利要求书限定。

Claims (30)

  1. 一种基于问答的搜索结果调整方法，包括：
    接收来自用户的搜索词，分析所述搜索词是否为问答类搜索词；
    当所述搜索词为问答类搜索词时，在问答知识库中匹配所述搜索词对应的问答类结果；
    利用所述问答类结果对搜索结果进行调整，将调整后的搜索结果提供给所述用户。
  2. 根据权利要求1所述的方法，其中，利用所述问答类结果对搜索结果进行调整，包括：
    利用所述问答类结果对根据所述搜索词在网页知识库中搜索得到的自然搜索结果进行调整。
  3. 根据权利要求2所述的方法，其中，利用所述问答类结果对根据所述搜索词在网页知识库中搜索得到的自然搜索结果进行调整，包括：
    判断所述自然搜索结果中与所述搜索词的相似度小于或等于预设阈值的搜索条目；
    将与所述搜索词的相似度小于或等于预设阈值的搜索条目替换为所述问答类结果。
  4. 根据权利要求1-3任一项所述的方法，其中，若所述问答类结果中问答类条目的数量为多条，利用所述问答类结果对搜索结果进行调整之前，还包括：
    将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。
  5. 根据权利要求4所述的方法，其中，将来自同一问答知识库的多条问答类条目合并为一条整合问答结果，包括：
    对于来自同一问答知识库中的多条问答类条目，选取其中的一条或多条作为显式的问答结果；
    将其余的问答类条目转化为一搜索用的链接元素与所述显式的问答结果整合，生成一条整合问答结果，其中，所述链接元素带有向该问答知识库依据所述搜索词发起搜索请求的链接信息。
  6. 根据权利要求1-5任一项所述的方法，其中，分析所述搜索词是否为问答类搜索词，包括：
    在疑问词库中匹配所述搜索词，其中，所述疑问词库中存储了通过数据挖掘方式或人工配置方式得到的疑问词；
    若存在匹配所述搜索词的疑问词，则确定所述搜索词为问答类搜索词。
  7. 一种搜索结果整合方法，包括：
    接收来自用户的问答类搜索词，搜索得到与所述问答类搜索词匹配的多条搜索条目；
    提取所述多条搜索条目中的多条问答类条目；
    将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。
  8. 根据权利要求7所述的方法，其中，
    将来自同一问答知识库的多条问答类条目合并为一条整合问答结果之前，还包括：提取所述多条搜索条目中的多条问答类条目的资源地址数据；
    将来自同一问答知识库的多条问答类条目合并为一条整合问答结果，包括：基于所述资源地址数据将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。
  9. 根据权利要求7或8所述的方法，其中，将来自同一问答知识库的多条问答类条目整合为一条整合问答结果之后，还包括：
    利用所述整合问答结果以及所述多条搜索条目中未被整合的搜索条目绘制搜索结果页，并提供给所述用户。
  10. 根据权利要求7-9任一项所述的方法，其中，将来自同一问答知识库的多条问答类条目合并为一条整合问答结果，包括：
    对于来自同一问答知识库的多条问答类条目，选取其中的一条或多条作为显式的问答结果；
    将其余的问答类条目转化为一搜索用的链接元素与所述显式的问答结果合并，生成一条整合问答结果，其中，所述链接元素带有向该问答知识库依据所述搜索词发起搜索请求的链接信息。
  11. 根据权利要求7-10任一项所述的方法，其中，对于来自同一问答知识库的多条问答类条目，选取其中的一条或多条作为显式的问答结果，包括：
    基于所述同一问答知识库的多条问答类条目的用户点击量，选取其中的一条或多条作为显式的问答结果。
  12. 根据权利要求7-11任一项所述的方法，其中，还包括：
    在所述显式的问答结果中选取一条问答类条目作为主问答结果；
    将其余的问答类条目以从属问答结果的身份与所述主问答结果整合，生成显式的问答结果。
  13. 根据权利要求7-12任一项所述的方法，其中，还包括：
    展示所述主问答结果的部分或全部的答案信息，且以文字链接的方式展示所述从属问答结果。
  14. 根据权利要求7-13任一项所述的方法，其中，提取所述多条搜索条目中的多条问答类条目，包括：
    通过后台问答知识库提取所述多条搜索条目中的多条问答类条目，其中，所述问答知识库通过数据挖掘建立问答类数据条目；或者，
    通过所述多条搜索条目中是否存在表明该搜索条目为问答类条目的标识，提取所述多条搜索条目中的多条问答类条目。
  15. 一种基于问答的搜索结果调整装置，包括：
    数据输入器，适于接收来自用户的搜索词；
    分析器，适于分析所述搜索词是否为问答类搜索词；
    匹配器，适于当所述搜索词为问答类搜索词时，在问答知识库中匹配所述搜索词对应的问答类结果；
    调整器，适于利用所述问答类结果对搜索结果进行调整；
    数据输出器，适于将调整后的搜索结果提供给所述用户。
  16. 根据权利要求15所述的装置，其中，所述调整器还适于：
    利用所述问答类结果对根据所述搜索词在网页知识库中搜索得到的自然搜索结果进行调整。
  17. 根据权利要求16所述的装置，其中，所述调整器还适于：
    判断所述自然搜索结果中与所述搜索词的相似度小于或等于预设阈值的搜索条目；
    将与所述搜索词的相似度小于或等于预设阈值的搜索条目替换为所述问答类结果。
  18. 根据权利要求15-17任一项所述的装置，其中，若所述问答类结果中问答类条目的数量为多条，所述调整器利用所述问答类结果对搜索结果进行调整之前，还包括：
    整合器，适于将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。
  19. 根据权利要求18所述的装置，其中，所述整合器还适于：
    对于来自同一问答知识库中的多条问答类条目，选取其中的一条或多条作为显式的问答结果；
    将其余的问答类条目转化为一搜索用的链接元素与所述显式的问答结果整合，生成一条整合问答结果，其中，所述链接元素带有向该问答知识库依据所述搜索词发起搜索请求的链接信息。
  20. 根据权利要求15-19任一项所述的装置，其中，所述分析器还适于：
    在疑问词库中匹配所述搜索词，其中，所述疑问词库中存储了通过数据挖掘方式或人工配置方式得到的疑问词；
    若存在匹配所述搜索词的疑问词，则确定所述搜索词为问答类搜索词。
  21. 一种搜索结果整合装置，包括：
    数据输入器，适于接收来自用户的问答类搜索词；
    搜索器，适于搜索得到与所述问答类搜索词匹配的多条搜索条目；
    提取器，适于提取所述多条搜索条目中的多条问答类条目；
    整合器，适于将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。
  22. 根据权利要求21所述的装置，其中，
    所述提取器还适于：提取所述多条搜索条目中的多条问答类条目的资源地址数据；
    所述整合器还适于：基于所述资源地址数据将来自同一问答知识库的多条问答类条目合并为一条整合问答结果。
  23. 根据权利要求21或22所述的装置，其中，所述整合器将来自同一问答知识库的多条问答类条目合并为一条整合问答结果之后，还包括：
    数据输出器，适于利用所述整合问答结果以及所述多条搜索条目中未被整合的搜索条目绘制搜索结果页，并提供给所述用户。
  24. 根据权利要求21-23任一项所述的装置，其中，所述整合器还适于：
    对于来自同一问答知识库的多条问答类条目，选取其中的一条或多条作为显式的问答结果；
    将其余的问答类条目转化为一搜索用的链接元素与所述显式的问答结果合并，生成一条整合问答结果，其中，所述链接元素带有向该问答知识库依据所述搜索词发起搜索请求的链接信息。
  25. 根据权利要求21-24任一项所述的装置，其中，所述整合器还适于：
    基于所述同一问答知识库的多条问答类条目的用户点击量，选取其中的一条或多条作为显式的问答结果。
  26. 根据权利要求21-25任一项所述的装置，其中，所述整合器还适于：
    在所述显式的问答结果中选取一条问答类条目作为主问答结果；
    将其余的问答类条目以从属问答结果的身份与所述主问答结果合并，生成显式的问答结果。
  27. 根据权利要求21-26任一项所述的装置，其中，所述整合器还适于：
    展示所述主问答结果的部分或全部的答案信息，且以文字链接的方式展示所述从属问答结果。
  28. 根据权利要求21-27任一项所述的装置，其中，所述提取器还适于：
    通过后台问答知识库提取所述多条搜索条目中的多条问答类条目，其中，所述问答知识库通过数据挖掘建立问答类数据条目；或者，
    通过所述多条搜索条目中是否存在表明该搜索条目为问答类条目的标识，提取所述多条搜索条目中的多条问答类条目。
  29. 一种计算机程序，包括计算机可读代码，当所述计算机可读代码在计算设备上运行时，导致所述计算设备执行根据权利要求1至14任一项所述的方法。
  30. 一种计算机可读介质，其中存储了如权利要求29所述的计算机程序。
PCT/CN2015/094619 2014-12-23 2015-11-13 基于问答的搜索结果调整方法和装置 WO2016101727A1 (zh)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN201410816126.6 2014-12-23
CN201410816126.6A CN105786873B (zh) 2014-12-23 2014-12-23 基于问答的搜索结果调整方法和装置
CN201410812104.2A CN105786852B (zh) 2014-12-23 2014-12-23 搜索结果整合方法和装置
CN201410812104.2 2014-12-23

Publications (1)

Publication Number Publication Date
WO2016101727A1 true WO2016101727A1 (zh) 2016-06-30

Family

ID=56149204

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/094619 WO2016101727A1 (zh) 2014-12-23 2015-11-13 基于问答的搜索结果调整方法和装置

Country Status (1)

Country Link
WO (1) WO2016101727A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109271495A (zh) * 2018-08-14 2019-01-25 阿里巴巴集团控股有限公司 问答识别效果检测方法、装置、设备及可读存储介质
CN110263137A (zh) * 2019-05-31 2019-09-20 阿里巴巴集团控股有限公司 主题关键词的提取方法和装置、电子设备
CN112101035A (zh) * 2020-09-21 2020-12-18 贝壳技术有限公司 命名实体识别方法、装置、可读存储介质及电子设备

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102375823A (zh) * 2010-08-13 2012-03-14 腾讯科技（深圳）有限公司 搜索结果聚合显示方法及***
CN104050286A (zh) * 2014-06-27 2014-09-17 北京奇虎科技有限公司 一种提供搜索结果整合的方法和装置

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102375823A (zh) * 2010-08-13 2012-03-14 腾讯科技（深圳）有限公司 搜索结果聚合显示方法及***
CN104050286A (zh) * 2014-06-27 2014-09-17 北京奇虎科技有限公司 一种提供搜索结果整合的方法和装置

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109271495A (zh) * 2018-08-14 2019-01-25 阿里巴巴集团控股有限公司 问答识别效果检测方法、装置、设备及可读存储介质
CN110263137A (zh) * 2019-05-31 2019-09-20 阿里巴巴集团控股有限公司 主题关键词的提取方法和装置、电子设备
CN110263137B (zh) * 2019-05-31 2023-06-06 创新先进技术有限公司 主题关键词的提取方法和装置、电子设备
CN112101035A (zh) * 2020-09-21 2020-12-18 贝壳技术有限公司 命名实体识别方法、装置、可读存储介质及电子设备
CN112101035B (zh) * 2020-09-21 2022-08-09 贝壳技术有限公司 命名实体识别方法、装置、可读存储介质及电子设备

Similar Documents

Publication Publication Date Title
US10146862B2 (en) Context-based metadata generation and automatic annotation of electronic media in a computer network
US11797626B2 (en) Search result filters from resource content
US9305083B2 (en) Author disambiguation
US9336318B2 (en) Rich content for query answers
US10977317B2 (en) Search result displaying method and apparatus
US20170255652A1 (en) Method for dynamically matching images with content items based on keywords in response to search queries
US20160034514A1 (en) Providing search results based on an identified user interest and relevance matching
US9235643B2 (en) Method and system for generating search results from a user-selected area
US20150052125A1 (en) Uniform search, navigation and combination of heterogeneous data
US10832049B2 (en) Electronic document classification system optimized for combining a plurality of contemporaneously scanned documents
CN105095440A (zh) 一种搜索推荐方法及装置
WO2016101716A1 (zh) 基于用户搜索意图的搜索方法及装置
WO2015081792A1 (zh) 联动性和个性化扩展搜索方法及装置、***
US10635705B2 (en) Methods, systems and computer readable storage media for determining relevant documents based on citation information
WO2015188719A1 (zh) 结构化数据与图片的关联方法与关联装置
US11816135B2 (en) Method, system, and computer program product for retrieving relevant documents
WO2016101727A1 (zh) 基于问答的搜索结果调整方法和装置
WO2015096625A1 (zh) 一种信息碎片的翻译方法及***
US20120179709A1 (en) Apparatus, method and program product for searching document
US20090187551A1 (en) Search results when searching for records of a business object
CN106919593B (zh) 一种搜索的方法和装置
US9990444B2 (en) Apparatus and method for supporting visualization of connection relationship
CN106649750B (zh) 针对多义项词条的搜索方法及装置
CN104462552A (zh) 问答页面核心词提取方法和装置
WO2015143911A1 (zh) 推送包含时效性信息的网页的方法和装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15871797

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15871797

Country of ref document: EP

Kind code of ref document: A1