CN110941695A - Question and answer information acquisition method and device, electronic equipment and storage medium - Google Patents
Question and answer information acquisition method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN110941695A CN110941695A CN201911075566.XA CN201911075566A CN110941695A CN 110941695 A CN110941695 A CN 110941695A CN 201911075566 A CN201911075566 A CN 201911075566A CN 110941695 A CN110941695 A CN 110941695A
- Authority
- CN
- China
- Prior art keywords
- question
- answer information
- target
- answer
- keyword
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 23
- 238000010586 diagram Methods 0.000 description 10
- 230000004044 response Effects 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Human Computer Interaction (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application relates to a question and answer information acquisition method, a question and answer information acquisition device, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring related words of the high-frequency problem keywords from a related word library; acquiring at least one candidate answer information based on the high-frequency question key words and the associated words of the high-frequency question key words, and determining at least one target answer information based on the at least one candidate answer information; at least one first question-answering message is generated. When corresponding question and answer information is generated for the non-answer question sentences, the semantics of each non-answer question sentence are not required to be analyzed one by one in a manual mode for each non-answer question sentence, the answer information corresponding to the non-answer question sentences is input in a manual mode to generate the corresponding question and answer information, the cost for generating the question and answer information is reduced, and the efficiency for generating the question and answer information is improved.
Description
Technical Field
The disclosure relates to the field of artificial intelligence, in particular to a question and answer information acquisition method and device, electronic equipment and a storage medium.
Background
The question-answering robot is an intelligent application which automatically feeds back answer information according to questions input by a user. The basis of the automatic feedback of the answer information by the question-answering robot is the question-answering information comprising question sentences and answer information corresponding to the question sentences in the knowledge base.
In the case where the sentence input by the user is, for example, a new type of question sentence, answer information cannot be fed back because answer information corresponding to the new type of question sentence does not exist in the knowledge base. The question sentence which does not correspond to the answer information becomes a question sentence without answer. Therefore, in order to enable the question-answering robot to feed back answer information for the question sentences without answers, it is necessary to periodically configure answer information corresponding to the question sentences without answers, generate question-answering information including the question sentences without answers and the answer information corresponding to the question sentences without answers, and add the generated question-answering information to the knowledge base.
At present, the commonly adopted method for generating question and answer information including answer-free question sentences and answer-free question sentences is as follows: and aiming at each statement without answer questions, analyzing the semantics of each statement without answer questions one by one in a manual mode, and inputting answer information corresponding to the statement without answer questions one by one in a manual mode.
However, the number of the non-answer question sentences may be in a large number, and for each non-answer question sentence, the semantics of each non-answer question sentence need to be analyzed one by one in a manual manner, and the answer information corresponding to each non-answer question sentence is input one by one, which results in extremely high labor cost and low efficiency.
Disclosure of Invention
In order to overcome the problems in the related art, the present disclosure provides a method and an apparatus for obtaining question and answer information, an electronic device, and a storage medium.
According to a first aspect of the embodiments of the present disclosure, there is provided a method for acquiring question and answer information, including:
acquiring related words of high-frequency question keywords from a related word library, wherein the high-frequency question keywords are keywords commonly included by a plurality of questions without answers, and the questions without answers are questions which do not correspond to answer information;
acquiring at least one candidate answer information based on the high-frequency question keyword and the associated words of the high-frequency question keyword, and determining at least one target answer information based on the at least one candidate answer information, wherein the target answer information corresponds to at least one non-answer question sentence in the plurality of non-answer question sentences;
generating at least one first question-answer message, wherein the first question-answer message comprises: a target answer information, one of the no answer question sentences corresponding to the target answer information.
According to a second aspect of the embodiments of the present disclosure, there is provided a question-answer information acquisition apparatus including:
a related word obtaining unit configured to obtain a related word of a high-frequency question keyword from a related word library, wherein the high-frequency question keyword is a keyword commonly included in a plurality of non-answer question sentences, and the non-answer question sentences are question sentences not corresponding to answer information;
an answer information determination unit configured to acquire at least one candidate answer information based on the high-frequency question keyword and the associated word of the high-frequency question keyword, and determine at least one target answer information based on the at least one candidate answer information, wherein the target answer information corresponds to at least one non-answer question sentence among the plurality of non-answer question sentences;
a first question-and-answer information generating unit configured to generate at least one first question-and-answer information including: a target answer information, one of the no answer question sentences corresponding to the target answer information.
The technical scheme provided by the embodiment of the disclosure can have the following beneficial effects:
when generating corresponding question and answer information for a question sentence without answer, candidate answer information possibly corresponding to the question sentence without answer is acquired, and target answer information corresponding to the question sentence without answer is obtained based on the acquired candidate answer information, for example, by directly selecting the target answer information corresponding to the question sentence without answer from all the candidate information, so as to generate the question and answer information. Therefore, the semantics of each non-answer question sentence are not required to be analyzed one by one in a manual mode aiming at each non-answer question sentence, the answer information corresponding to the non-answer question sentence is input in a manual mode to generate the corresponding question and answer information, the cost for generating the question and answer information is reduced, and the efficiency for generating the question and answer information is improved.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and together with the description, serve to explain the principles of the disclosure.
Fig. 1 is a flowchart illustrating one method for acquiring question answering information according to an embodiment of the present disclosure;
FIG. 2 is a diagram illustrating an effect of an information presentation interface;
fig. 3 is a schematic diagram illustrating a principle of corresponding a related question sentence with target answer information;
fig. 4 is a schematic structural diagram illustrating a question answering information acquiring apparatus according to an embodiment of the present application.
Detailed Description
The present disclosure is described in further detail below with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the relevant invention and not restrictive of the invention. It should be noted that, for convenience of description, only the portions related to the related invention are shown in the drawings.
It should be noted that, in the present disclosure, the embodiments and features of the embodiments may be combined with each other without conflict. The present disclosure will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.
Fig. 1 is a flowchart of a method for acquiring question and answer information according to an embodiment of the present disclosure. The method comprises the following steps:
In the present application, a question sentence without answer is a question sentence without answer information. After receiving the sentence input by the user each time, the question-answering robot determines that the knowledge base does not include answer information corresponding to the sentence input by the user, and then the sentence input by the user can be used as a no-answer question sentence.
In the present application, for one high-frequency question keyword, the high-frequency question keyword is commonly included by a plurality of unanswered question sentences.
In the present application, each non-answer question sentence may be segmented to obtain a plurality of terms in the non-answer question sentence, terms with a low association degree with the query destination, such as "what" and "what" in each non-answer question sentence, are removed, and the keywords with a high association degree with the query destination in each non-answer question sentence are all added to the non-answer question keyword library. Then, each of the keywords included in common in the plurality of non-answer question sentences may be determined from the non-answer question keyword library, and each of the keywords included in common in the plurality of non-answer question sentences may be regarded as a high-frequency question keyword.
In the present application, the associated word library stores an associated word for each of a plurality of words and a plurality of words. For each term in the associated term library, the term may have one or more associated terms. For each word, the associated word of the word may be a synonym, a near-synonym, a word that can form a phrase with the word, and the like.
In the present application, for a high-frequency question keyword appearing in a plurality of unanswered question sentences, when obtaining a related term of the high-frequency question keyword from a related term library, the high-frequency question keyword may be first found in the related term library, and then one or more related terms of the high-frequency question keyword may be found in the related term library.
102, obtaining at least one candidate answer information based on the high-frequency question key words and the associated words of the high-frequency question key words, and determining at least one target answer information based on the at least one candidate answer information.
In the present application, for a high-frequency question keyword, after obtaining a related word of the high-frequency question keyword from a related word library, at least one candidate answer information may be obtained based on the high-frequency question keyword and the related word of the high-frequency question keyword, and at least one target answer information may be determined based on the at least one candidate answer information. The determined target answer information corresponds to at least one non-answer question sentence among a plurality of non-answer question sentences including the high-frequency question keyword.
In the application, for a high-frequency question keyword, when at least one candidate answer information is obtained based on the high-frequency question keyword and the associated word of the high-frequency question keyword, the answer information of the associated word including the high-frequency question keyword and/or the answer information of the associated word including the high-frequency question keyword can be found from a knowledge base, and each found answer information can be respectively used as one candidate answer information.
In the present application, the number of the determined target answer information may be plural for one high frequency question keyword. The target answer information may be candidate answer information corresponding to at least one of a plurality of answer-free question sentences including the high-frequency question keyword among all the candidate answer information. When a candidate answer information may correspond to at least one non-answer question sentence among a plurality of non-answer question sentences including the high-frequency question keyword, the candidate answer information may be a target answer information.
In the application, for a high-frequency question keyword, all answer-free question sentences including the high-frequency question keyword and all acquired candidate answer information can be displayed in an information display interface. The correspondence of the presented answer question sentence with the presented candidate answer information may be determined by the user participating in the generation of the question-and-answer information. When the user participating in the generation of the question-and-answer information determines that one candidate answer information to be presented may correspond to one presented non-answer question sentence including the high-frequency question keyword, the user participating in the generation of the question-and-answer information may select the one candidate answer information as a target answer information in the information presentation interface, and at the same time, the user participating in the generation of the question-and-answer information may select the one non-answer question sentence including the high-frequency question keyword in the information presentation interface. Then, when the user participating in the generation of the question-answer information performs an operation of instructing to correspond the non-answer question sentence with the corresponding answer information in the information presentation interface, the target answer information may be used as the target answer information corresponding to the non-answer question sentence including the high-frequency question keyword.
Referring to the above-described manner of obtaining target answer information corresponding to a question-and-answer-free sentence, whenever a candidate answer information determined to be presented by a user participating in the generation of question-and-answer information may be used as a target answer information corresponding to a question-and-answer-free sentence including the high-frequency question keyword, target answer information corresponding to a question-and-answer-free sentence including the high-frequency question keyword may be determined.
In the application, after obtaining all candidate answer information for a high-frequency question keyword, a user participating in generating question-answer information instructs to perform corresponding operation on a question-free question sentence and corresponding answer information in an information display interface, and one candidate answer information can be used as target answer information commonly corresponding to a plurality of question-free sentences including the high-frequency question keyword. For a question sentence without answer including the high-frequency question keyword, the question sentence without answer including the high-frequency question keyword may correspond to one or more target answer information by a user participating in generating question-answer information instructing to correspond the question sentence without answer to the corresponding answer information in an information presentation interface.
In some embodiments, obtaining at least one candidate answer information based on the high-frequency question keyword and the associated word of the high-frequency question keyword, and determining at least one target answer information based on the at least one candidate answer information includes: when matching answer information exists in the knowledge base, the matching answer information is used as candidate answer information, wherein the matching answer information is the answer information of associated words at least comprising the high-frequency question keywords in the knowledge base; when matching answer information does not exist in the knowledge base, preset answer information is used as candidate answer information; displaying the high-frequency question key words, the associated words of the high-frequency question key words and all candidate answer information; and determining candidate answer information selected by the target user from all the presented candidate answer information or answer information input by the target user based on the presented candidate answer information as the target answer information.
In the present application, for a high-frequency question keyword, when candidate answer information is obtained based on the high-frequency question keyword and a related word of the high-frequency question keyword, it may be first determined whether there is matching answer information in a knowledge base. And matching answer information is answer information of at least the associated words of the high-frequency question keywords in the knowledge base. Answer information including at least the associated word of the high-frequency question keyword as matching answer information may further include the high-frequency question keyword.
In the present application, when there is matching answer information in the knowledge base, the number of matching answer information in the knowledge base may be one or more. And each piece of matching answer information is taken as candidate answer information.
In the present application, when candidate answer information does not exist in the knowledge base, the preset answer information may be used as the candidate answer information. The number of the preset answer information may be plural. Each preset answer information is one of the following: answer information added into the knowledge base in a preset time period and answer information with the feedback times exceeding a time threshold value in the knowledge base in the preset time period. The preset time period may be a time period in which a date having a length from the current date smaller than a preset number of days is used as a start date and the current date is used as an end date.
In the present application, for a high-frequency question keyword, after obtaining all candidate answer information, the high-frequency question keyword, the associated word of the high-frequency question keyword, and all the obtained candidate answer information may be displayed in an information display interface. Meanwhile, a plurality of non-answer question sentences including the high-frequency question keywords may be presented in an information presentation interface.
In some embodiments, presenting the high frequency question keyword and the associated words of the high frequency question keyword and the candidate answer information comprises: and displaying the high-frequency question keywords and the associated words of the high-frequency question keywords in a map form.
In the application, for a high-frequency question keyword, the high-frequency question keyword and the associated words of the high-frequency question keyword can be displayed in an information display interface in a map form. When the high-frequency problem keyword and the associated terms of the high-frequency problem keyword are displayed in the information display interface in a graph form, the high-frequency problem keyword and each associated term of the high-frequency problem keyword can be represented by one node, and each associated term node representing the high-frequency problem keyword is connected with the node representing the high-frequency problem keyword.
In the present application, the target user may be a person participating in generating answer information, such as an engineer participating in generating answer information.
In the present application, the number of the determined target answer information may be one or more for one high frequency question keyword. The target answer information may be selected by the target user from all candidate answer information. The target answer information may also be answer information input by the target user based on the candidate answer information. Each target answer information is one of the following: the target user selects one candidate answer information from all candidate answer information and the answer information input by the target user. The answer information input by the target user can be obtained by modifying the candidate answer information.
In the present application, for a high frequency question keyword, when a target user selects a candidate answer information as a target answer information from all the presented candidate answer information, the target user may select one or more non-answer question sentences including the high frequency question keyword corresponding to the target answer information from among all non-answer question sentences including the high frequency question keyword while the target user selects the target answer information. Then, the target user may perform an operation of instructing to correspond the non-answer question sentence including the high-frequency question keyword with the corresponding answer information, for example, click a button on the information presentation interface instructing to correspond the non-answer question sentence including the high-frequency question keyword with the corresponding answer information. The target answer information may be corresponded to one or more non-answer question sentences including the high-frequency question keyword selected by the target user in response to an operation of the target user indicating to correspond the non-answer question sentences including the high-frequency question keyword with the corresponding answer information. Accordingly, the target answer information serves as the target answer information corresponding to the one or more non-answer question sentences including the high-frequency question keyword selected by the user.
In the present application, when a target user inputs answer information based on one candidate answer information, for example, when the answer information input by the user is obtained by modifying the candidate answer information, the answer information input by the user may be used as one target answer information, and the target user may select one or more non-answer question sentences including the high-frequency question keyword corresponding to the target answer information among a plurality of non-answer question sentences including the high-frequency question keyword while the target answer information is input by the user. Then, the target user may perform an operation indicating that the no-answer question sentence including the high-frequency question keyword is to be corresponded with the corresponding answer information. The target answer information may be corresponded to one or more non-answer question sentences including the high-frequency question keyword selected by the target user in response to an operation of the target user to correspond the non-answer question sentences including the high-frequency question keyword to the corresponding answer information.
For example, a plurality of question sentences each associated with an object, proton, include "proton", and the knowledge base has no answer information associated with proton. The plurality of question sentences including "protons" are a plurality of answer-free question sentences not corresponding to answer information, and the "protons" are high-frequency question keywords. After presenting all the candidate answer information, the target user determines that each of the candidate answer information of all the candidate answer information cannot correspond to the no answer question sentence including "proton". Answer information may be input by the target user. The answer information input by the target user includes insurance clause information, product introduction information, etc. related to the protons. The answer information input by the target user may be one target answer information. While the user inputs answer information, the target user may select one or more of a plurality of answer-free question sentences including "protons". Then, the target user may perform an operation indicating that the no-answer question sentence including the high-frequency question keyword is to be corresponded with the corresponding answer information, may correspond the no-answer question sentence including the high-frequency question keyword with the corresponding answer information in response to the indication of the target user, and may correspond the target answer information input by the target user with one or more of the plurality of no-answer question sentences including "proton".
In the present application, a first question and answer message is composed of a target answer message and a question sentence without answer including a high frequency question keyword corresponding to the target answer message. For one target answer information, when the number of answer-free question sentences including high-frequency question keywords corresponding to the target answer information is plural, the target answer information may be combined with each of the answer-free question sentences including high-frequency question keywords corresponding to the target answer information, respectively, to obtain a plurality of first question-answer information, each of the obtained plurality of first question-answer information including the target answer information and one answer-free question sentence including high-frequency question keywords corresponding to the target answer information.
In the present application, after generating the at least one first question and answer information, the at least one question and answer information may be added to the knowledge base. Thus, the non-answer question sentence in the at least one first question-and-answer information becomes an answer question sentence corresponding to the at least one target answer information. After the at least one question-answer information is added into the knowledge base, when the question-answer robot receives the question sentences without answer in the at least one first question-answer information or the question sentences with similar semanteme to the question sentences without answer in the at least one first question-answer information input by the user again, the corresponding answer information can be found according to the at least one first question-answer information, and the found answer information is fed back to the user.
In some embodiments, further comprising: generating at least one suggested question sentence based on the high-frequency question keywords, the associated words of the high-frequency question keywords and the keywords in the target answer information; determining at least one target question sentence based on the at least one suggested question sentence, wherein the target question sentence corresponds to the at least one target answer information; generating at least one second question-answer message, wherein the second question-answer message comprises: a target question sentence, and target answer information corresponding to the target question sentence.
In the present application, for one high frequency question keyword, when at least one suggested question sentence is generated based on the high frequency question keyword and associated words of the high frequency question keyword and keywords in the target answer information, the target answer information including the high frequency question keyword and/or associated words of the high frequency question keyword in all the target answer information may be determined. Then, other keywords than the high-frequency question keyword and the associated words of the high-frequency question keyword in the target answer information including the high-frequency question keyword and/or the associated words of the high-frequency question keyword may be determined. All other keywords may be combined with the high frequency question keywords and/or associated terms of the high frequency question keywords to arrive at one or more suggested question sentences.
In the present application, the number of target question sentences may be one or more. One target sentence may correspond to one or more target answer information. When at least one target question sentence is determined based on at least one suggested question sentence, the at least one suggested question sentence may be presented on the information presentation interface, and then one or more target question sentences may be selected from the at least one suggested question sentence by the target user, thereby determining the one or more target question sentences.
In the present application, for one target question sentence, while the target user selects the target question sentence, the target user may select one or more target answer information corresponding to the target question sentence among all the target answer information. Then, the target user may perform an operation of instructing to correspond the target question sentence to the target answer information, for example, click a button on the information presentation interface instructing to correspond the target question sentence to the target answer information. The target question sentence and the target answer information may be subjected to corresponding operation in response to an instruction of a target user, and one or more pieces of target answer information selected by the user may be used as one or more pieces of target answer information corresponding to the target question sentence, so that one or more pieces of target answer information corresponding to the target question sentence may be obtained.
Referring to the above-mentioned manner of obtaining one or more target answer information corresponding to one target question sentence, one or more target answer information corresponding to each target question sentence can be obtained respectively.
In this application, after obtaining one or more target answer information corresponding to each target question sentence, at least one second question-answer information may be generated. A second question and answer information is composed of a target question sentence and target answer information corresponding to the target question sentence. For a target question sentence, when the target question sentence corresponds to a plurality of target answer information, the target question sentence may be combined with one of the plurality of target answer information corresponding to the target question sentence, respectively, to obtain a plurality of second question-answer information, and the obtained plurality of second question-answer information includes the target question sentence and one of the target answer information corresponding to the target question sentence.
After generating the at least one second question-and-answer message, the at least one second question-and-answer message may be added to the knowledge base. After the at least one piece of second question-answering information is added into the knowledge base, when the question-answering robot receives a target question sentence or a question sentence similar to the target question sentence input by the user, corresponding answer information can be found according to the at least one piece of second question-answering information, and the found answer information is fed back to the user.
In some embodiments, generating at least one suggested question sentence based on the high frequency question keyword and the associated terms of the high frequency question keyword and the keywords in the target answer information includes: determining at least one co-occurrence keyword, wherein the co-occurrence keyword is a word except for the associated word of the high-frequency question keyword, and the frequency of the co-occurrence keyword and the high-frequency question keyword appearing in the target answer information at the same time is greater than a threshold value; and generating at least one suggested question sentence based on the high-frequency question keyword, the associated words of the high-frequency question keyword and the at least one co-occurrence keyword.
In the present application, for a high-frequency question keyword, when one piece of target answer information of all the determined target answer information includes the high-frequency question keyword and a keyword other than the associated word of the high-frequency question keyword, the high-frequency question keyword and the keyword other than the associated word of the high-frequency question keyword are simultaneously present in the target answer information, and the keyword other than the associated word of the high-frequency question keyword may be regarded as a co-occurrence keyword. For one co-occurrence keyword, the number of times that the co-occurrence keyword and the high-frequency question keyword simultaneously appear in the target answer information is the number of target answer information including the high-frequency question keyword and the co-occurrence keyword simultaneously in all the target answer information.
In the present application, for a high frequency question keyword, the determined at least one co-occurrence keyword may be combined with the high frequency question keyword and/or associated terms of the high frequency question keyword to obtain one or more suggested question sentences.
In some embodiments, at least one co-occurrence keyword is added to the associated word library, the co-occurrence keywords being associated words of the high frequency question keywords, respectively.
In the present application, for a high-frequency question keyword, after determining at least one co-occurrence keyword, each co-occurrence keyword may be added to the associated word library, and each co-occurrence keyword is respectively used as an associated word of the high-frequency question keyword.
In some embodiments, determining the target question statement based on the at least one suggested question statement comprises: and determining the suggested question sentence selected by the target user from the at least one suggested question sentence or the question sentence input by the target user as the target question sentence.
In the present application, the target question statement may be selected by the target user from all of the suggested question statements. The target question sentence may also be a target question sentence that may be input by the target user based on the suggested question sentence. The target question sentence input by the target user can be obtained by modifying the suggested question sentence. Each target question statement is one of: the target user selects one suggested question sentence from all suggested question sentences and the target user inputs one question sentence.
Please refer to fig. 2, which shows an effect diagram of the information display interface.
In FIG. 2, an information presentation interface 201 is shown. Related words of the high-frequency problem keyword such as "naval" and "seventy-eight", "forty-five", "cast military" and the like displayed in the information display interface 201 in the form of a map. And displaying the suggested question sentences on the left side of the information display interface and displaying all target answer information on the right side of the information display interface. In addition, the no-answer question sentences including the high-frequency question keyword 'army' can be displayed in the information display interface at the same time. The keywords obtained from the target answer information may include "solution", "allowance". The high-frequency question keywords such as "admiration" and "seventy eight", "forty five" and the like are combined with the co-occurrence keywords such as "scheme", "allowance" and the like, so that the suggested question sentences can be obtained. Only information presented in the information presentation interface 201 when the suggested question sentence corresponds to the target answer information is shown in fig. 2. When the high-frequency question keyword 'nav' is corresponded to the corresponding answer information, the left side of the information display interface can display the no-answer question sentence including the high-frequency question keyword 'nav', and the right side of the information display interface can display all candidate answer information obtained based on the high-frequency question keyword 'nav' and the associated words of the high-frequency question keyword 'nav', such as 'seven eight', 'four five', 'cast military', and the like.
The target user can select the target question sentence from all the suggested question sentences by performing sentence selection operation in the information presentation interface. The target user can also input question sentences, for example, by modifying the suggested question sentences, using the sentences obtained after modification as the target question sentences, and displaying the question sentences input by the user in the information display interface.
When the target user selects one target question sentence, the target user can select one or more target answer information which can correspond to the target question sentence from all the target answer information by performing answer information selection operation in the information display interface. Then, the target user may perform an operation of instructing to correspond the target question sentence with the target answer information. The target question sentence and the target answer information may be subjected to corresponding operation in response to an instruction of a target user, and one or more pieces of target answer information selected by the user may be used as one or more pieces of target answer information corresponding to the target question sentence, so that one or more pieces of target answer information corresponding to the target question sentence may be obtained.
Please refer to fig. 3, which illustrates a schematic diagram of corresponding related question sentences and target answer information.
In the present application, a keyword library of no answer questions may be constructed according to a library of no answer questions. The answer-free question bank comprises a plurality of answer-free question sentences which do not correspond to the answer information. After receiving the sentence input by the user each time, the question-answering robot determines that the knowledge base does not include answer information corresponding to the sentence input by the user, and then the sentence input by the user can be used as a no-answer question sentence. Each non-answer question sentence can be participled to obtain a plurality of words in the non-answer question sentence. Terms with low relevance to the query purpose, such as "what is", in each non-answer question sentence, may be removed, and the keywords with high relevance to the query purpose in each non-answer question sentence are added to the non-answer question keyword library.
In the application, a high-frequency question keyword library can be constructed according to the answer-free question keyword library. High-frequency question keywords included in common by a plurality of answer-free question sentences may be determined from the answer-free question keyword library, and each of the determined high-frequency question keywords is added to the high-frequency question keyword library.
In the present application, for a high-frequency question keyword, the high-frequency question keyword may be obtained from a high-frequency question keyword library, and at the same time, a related term of the high-frequency question keyword may be obtained from a related term library. The related words of the high-frequency problem keyword can be synonyms of the high-frequency problem keyword, words which are preset and can form phrases with the high-frequency problem keyword, and the like.
In the present application, for a high-frequency question keyword, all candidate answer information may be obtained based on the high-frequency question keyword and associated words of the high-frequency question keyword. Then, at least one piece of target answer information may be determined further based on the candidate answer information. Each target answer information corresponds to at least one no-answer question sentence including the high-frequency question keyword.
In this application, for one high-frequency question keyword, after determining at least one piece of target answer information, at least one co-occurrence keyword may be further obtained from the target answer information, and the at least one co-occurrence keyword is combined with the high-frequency question keyword and/or associated words of the high-frequency question keyword to obtain one or more suggested question sentences. At the same time, the target user may also input question statements based on the suggested question statements.
In the present application, each of the related question sentences may be one of a no answer question sentence including a high frequency question keyword, a generated suggested question sentence, and a question sentence input by the target user based on the suggested question sentence. In the present application, at least part of the related question sentences may correspond to the corresponding target answer information. When the target user determines that one piece of target answer information may correspond to at least one no-answer question sentence, the one piece of target answer information may be respectively used as the target answer information corresponding to each no-answer question sentence in the at least one no-answer question sentence. When the target user determines that one suggested question sentence or one question sentence input by the target user can correspond to at least one piece of target answer information, each piece of target answer information in the at least one piece of target answer information can be used as one piece of target answer information corresponding to the one suggested question sentence or the one question sentence input by the target user respectively.
In the application, the co-occurrence keywords obtained from the target answer information may be added to the related term library, and in the related term library, the co-occurrence keywords obtained from the target answer information may be used as related terms of the high-frequency question keywords.
Please refer to fig. 4, which illustrates a schematic structural diagram of a question answering information obtaining apparatus according to an embodiment of the present application. The specific implementation of the operation that each unit in the question answering information acquisition device provided in the embodiment of the present application is configured to complete may refer to the specific implementation of the corresponding operation described in the method embodiment.
As shown in fig. 4, the question-answering information acquisition apparatus includes: a related word acquiring unit 401, an answer information determining unit 402, and a first question and answer information generating unit 403.
The related word obtaining unit 401 is configured to obtain a related word of a high-frequency question keyword from a related word library, where the high-frequency question keyword is a keyword commonly included in a plurality of answer-free question sentences, and the answer-free question sentences are question sentences not corresponding to answer information;
the answer information determination unit 402 is configured to obtain at least one candidate answer information based on the high-frequency question keyword and the associated word of the high-frequency question keyword, and determine at least one target answer information based on the at least one candidate answer information, wherein the target answer information corresponds to at least one answer-free question sentence among the plurality of answer-free question sentences;
the first question and answer information generating unit 403 is configured to generate at least one first question and answer information,
the first question-answering information includes: a target answer information, one of the no answer question sentences corresponding to the target answer information.
In some embodiments, the answer information determination unit 402 is further configured to: when matching answer information exists in the knowledge base, the matching answer information is used as candidate answer information, wherein the matching answer information is the answer information of associated words at least comprising the high-frequency question keywords in the knowledge base; when matching answer information does not exist in the knowledge base, preset answer information is used as candidate answer information; displaying the high-frequency question keywords, the associated words of the high-frequency question keywords and all candidate answer information; and determining candidate answer information selected by the target user from all the presented candidate answer information or answer information input by the target user based on the presented candidate answer information as the target answer information.
In some embodiments, presenting the high-frequency question keyword and the associated words of the high-frequency question keyword and all candidate answer information comprises: and displaying the high-frequency question keywords and the associated words of the high-frequency question keywords in a map form.
In some embodiments, the question-answer information acquisition means further includes: a second question and answer information generating unit configured to generate at least one suggested question sentence based on the high-frequency question keyword and the associated word of the high-frequency question keyword and a keyword in the target answer information; determining at least one target question sentence based on the at least one suggested question sentence, wherein the target question sentence corresponds to the at least one target answer information; generating at least one second question-answer message, wherein the second question-answer message comprises: a target question sentence, and one of the target answer information corresponding to the target question sentence.
In some embodiments, the second question and answer information generating unit is further configured to: determining at least one co-occurrence keyword, wherein the co-occurrence keyword is a keyword except associated words of the high-frequency question keyword, and the frequency of the co-occurrence keyword and the high-frequency question keyword appearing in the target answer information at the same time is greater than a threshold value; and generating at least one suggested question sentence based on the high-frequency question keyword, the associated word of the high-frequency question keyword and at least one co-occurrence keyword.
In some embodiments, the question-answer information acquisition means further includes: a related word adding unit configured to add at least one co-occurrence keyword to a related word library, wherein the co-occurrence keyword is a related word of the high-frequency question keyword.
In some embodiments, the second question and answer information generating unit is further configured to: displaying at least one suggested question sentence; and determining the suggested question sentence selected by the target user from the at least one suggested question sentence or the question sentence input by the target user as the target question sentence.
The present application further provides an electronic device that may be configured with one or more processors; a memory for storing one or more programs, the one or more programs may include instructions for performing the operations described in the above embodiments. The one or more programs, when executed by the one or more processors, cause the one or more processors to perform the instructions of the operations described in the above embodiments.
The present application also provides a computer readable medium, which may be included in an electronic device; or the device can be independently arranged and not assembled into the electronic equipment. The computer-readable medium carries one or more programs which, when executed by an electronic device, cause the electronic device to perform the operations described in the embodiments above.
It should be noted that the computer readable medium described herein can be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may include, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with a message execution system, apparatus, or device. In this application, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with a message execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable messages for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer messages.
The above description is only a preferred embodiment of the present request and is illustrative of the principles of the technology employed. It will be understood by those skilled in the art that the scope of the invention herein referred to is not limited to the technical embodiments with the specific combination of the above technical features, but also encompasses other technical embodiments with any combination of the above technical features or their equivalents without departing from the inventive concept. For example, technical embodiments formed by mutually replacing the above-mentioned features with (but not limited to) technical features having similar functions disclosed in the present application.
Claims (10)
1. A question-answer information acquisition method is characterized by comprising the following steps:
acquiring related words of high-frequency question keywords from a related word library, wherein the high-frequency question keywords are keywords commonly included by a plurality of questions without answers, and the questions without answers are questions which do not correspond to answer information;
acquiring at least one candidate answer information based on the high-frequency question keyword and the associated words of the high-frequency question keyword, and determining at least one target answer information based on the at least one candidate answer information, wherein the target answer information corresponds to at least one non-answer question sentence in the plurality of non-answer question sentences;
generating at least one first question-answer message, wherein the first question-answer message comprises: a target answer information, one of the no answer question sentences corresponding to the target answer information.
2. The method of claim 1, wherein obtaining at least one candidate answer information based on the high-frequency question keyword and the associated word of the high-frequency question keyword, and determining at least one target answer information based on the at least one candidate answer information comprises:
when matching answer information exists in the knowledge base, the matching answer information is used as candidate answer information, wherein the matching answer information is the answer information of associated words at least comprising the high-frequency question keywords in the knowledge base;
when matching answer information does not exist in the knowledge base, preset answer information is used as candidate answer information;
displaying the high-frequency question keywords, the associated words of the high-frequency question keywords and all candidate answer information;
and determining candidate answer information selected by the target user from all the presented candidate answer information or answer information input by the target user based on the presented candidate answer information as the target answer information.
3. The method of claim 2, wherein presenting the high frequency question keyword and associated words of the high frequency question keyword and all candidate answer information comprises:
and displaying the high-frequency question keywords and the associated words of the high-frequency question keywords in a map form.
4. The method according to one of claims 1 to 3, characterized in that the method further comprises:
generating at least one suggested question sentence based on the high-frequency question keywords, the associated words of the high-frequency question keywords and the keywords in the target answer information;
determining at least one target question sentence based on the at least one suggested question sentence, wherein the target question sentence corresponds to the at least one target answer information;
generating at least one second question-answer message, wherein the second question-answer message comprises: a target question sentence, and one of the target answer information corresponding to the target question sentence.
5. The method of claim 4, wherein generating at least one suggested question sentence based on the high frequency question keyword and associated words of the high frequency question keyword and keywords in the target answer information comprises:
determining at least one co-occurrence keyword, wherein the co-occurrence keyword is a keyword except associated words of the high-frequency question keyword, and the frequency of the co-occurrence keyword and the high-frequency question keyword appearing in the target answer information at the same time is greater than a threshold value;
and generating at least one suggested question sentence based on the high-frequency question keyword, the associated word of the high-frequency question keyword and at least one co-occurrence keyword.
6. The method of claim 5, further comprising:
and adding at least one co-occurrence keyword into a related word library, wherein the co-occurrence keyword is used as a related word of the high-frequency question keyword.
7. The method of claim 6, wherein determining at least one target question statement based on at least one suggested question statement comprises:
displaying at least one suggested question sentence;
and determining the suggested question sentence selected by the target user from the at least one suggested question sentence or the question sentence input by the target user as the target question sentence.
8. A question-answer information acquisition apparatus characterized by comprising:
a related word obtaining unit configured to obtain a related word of a high-frequency question keyword from a related word library, wherein the high-frequency question keyword is a keyword commonly included in a plurality of non-answer question sentences, and the non-answer question sentences are question sentences not corresponding to answer information;
an answer information determination unit configured to acquire at least one candidate answer information based on the high-frequency question keyword and the associated word of the high-frequency question keyword, and determine at least one target answer information based on the at least one candidate answer information, wherein the target answer information corresponds to at least one non-answer question sentence among the plurality of non-answer question sentences;
a first question-and-answer information generating unit configured to generate at least one first question-and-answer information including: a target answer information, one of the no answer question sentences corresponding to the target answer information.
9. An electronic device, comprising:
a processor;
a memory for storing the processor-executable instructions;
wherein the processor is configured to execute the instructions to implement the method of any one of claims 1 to 7.
10. A storage medium having instructions that, when executed by a processor of an electronic device, enable the electronic device to perform the method of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911075566.XA CN110941695A (en) | 2019-11-05 | 2019-11-05 | Question and answer information acquisition method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911075566.XA CN110941695A (en) | 2019-11-05 | 2019-11-05 | Question and answer information acquisition method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110941695A true CN110941695A (en) | 2020-03-31 |
Family
ID=69907455
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911075566.XA Pending CN110941695A (en) | 2019-11-05 | 2019-11-05 | Question and answer information acquisition method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110941695A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070118519A1 (en) * | 2005-11-21 | 2007-05-24 | Fuji Xerox Co., Ltd. | Question answering system, data search method, and computer program |
| WO2017016104A1 (en) * | 2015-07-28 | 2017-02-02 | 百度在线网络技术（北京）有限公司 | Question-answer information processing method and apparatus, storage medium, and device |
| CN108388558A (en) * | 2018-02-07 | 2018-08-10 | 平安普惠企业管理有限公司 | Problem matching process, device, customer service robot and storage medium |
| CN109271505A (en) * | 2018-11-12 | 2019-01-25 | 深圳智能思创科技有限公司 | A kind of question answering system implementation method based on problem answers pair |
| CN110019701A (en) * | 2017-09-18 | 2019-07-16 | 京东方科技集团股份有限公司 | Method, question and answer service system and storage medium for question and answer service |
| CN110209768A (en) * | 2019-05-31 | 2019-09-06 | 中国联合网络通信集团有限公司 | The problem of automatic question answering treating method and apparatus |
-
2019
- 2019-11-05 CN CN201911075566.XA patent/CN110941695A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070118519A1 (en) * | 2005-11-21 | 2007-05-24 | Fuji Xerox Co., Ltd. | Question answering system, data search method, and computer program |
| WO2017016104A1 (en) * | 2015-07-28 | 2017-02-02 | 百度在线网络技术（北京）有限公司 | Question-answer information processing method and apparatus, storage medium, and device |
| CN110019701A (en) * | 2017-09-18 | 2019-07-16 | 京东方科技集团股份有限公司 | Method, question and answer service system and storage medium for question and answer service |
| CN108388558A (en) * | 2018-02-07 | 2018-08-10 | 平安普惠企业管理有限公司 | Problem matching process, device, customer service robot and storage medium |
| CN109271505A (en) * | 2018-11-12 | 2019-01-25 | 深圳智能思创科技有限公司 | A kind of question answering system implementation method based on problem answers pair |
| CN110209768A (en) * | 2019-05-31 | 2019-09-06 | 中国联合网络通信集团有限公司 | The problem of automatic question answering treating method and apparatus |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210081725A1 (en) | Method, apparatus, server, and user terminal for constructing data processing model | |
| CN108170792B (en) | Question and answer guiding method and device based on artificial intelligence and computer equipment | |
| CN106649742B (en) | Database maintenance method and device | |
| KR101707369B1 (en) | Construction method and device for event repository | |
| CN109033244B (en) | Search result ordering method and device | |
| CN104657346A (en) | Question matching system and question matching system in intelligent interaction system | |
| CN111898643B (en) | Semantic matching method and device | |
| CN109635094B (en) | Method and device for generating answer | |
| US10664755B2 (en) | Searching method and system based on multi-round inputs, and terminal | |
| US10157203B2 (en) | Question transformation in question answer systems | |
| US20150234825A1 (en) | Question resolution processing in deep question answering systems | |
| EP2953038A1 (en) | Interactive searching method and apparatus | |
| CN109858045B (en) | Machine translation method and device | |
| US11151202B2 (en) | Exploiting answer key modification history for training a question and answering system | |
| US10795878B2 (en) | System and method for identifying answer key problems in a natural language question and answering system | |
| CN108519998B (en) | Problem guiding method and device based on knowledge graph | |
| CN104573099A (en) | Topic searching method and device | |
| CN108733712B (en) | Question-answering system evaluation method and device | |
| CN110008308B (en) | Method and device for supplementing information for user question | |
| CN110807311B (en) | Method and device for generating information | |
| CN116737908A (en) | Knowledge question-answering method, device, equipment and storage medium | |
| US20180067927A1 (en) | Customized Translation Comprehension | |
| CN112507139A (en) | Knowledge graph-based question-answering method, system, equipment and storage medium | |
| CN110232920B (en) | Voice processing method and device | |
| CN112148844A (en) | Information reply method and device for robot |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200331 |
|
| RJ01 | Rejection of invention patent application after publication |