Question and answer quality determination method, device, equipment and storage medium

Abstract

The present application discloses a method, device, equipment and storage medium for determining the quality of questions and answers, and relates to the technical field of natural language processing. The specific implementation scheme is: matching the question and answer pair to be identified with the candidate high-quality question and answer pair to obtain the target high-quality question and answer pair; wherein the question and answer pair to be identified includes the question to be identified and the answer to be identified, and the target high-quality question and answer pair includes the target high-quality question and the target high-quality answer; according to the first correlation between the answer to be identified and the target high-quality answer, and the second correlation between the question to be identified and the target high-quality question, the quality of the question and answer pair to be identified is determined. The present application can improve the accuracy of the question and answer quality.

Description

Technical Field

[0001] The present application relates to the field of Internet technology, in particular to the field of natural language processing technology, and specifically to a method, device, equipment and storage medium for determining the quality of questions and answers. Background Art

[0002] With the development of computer technology, question-and-answer Internet products have been increasingly widely used. In question-and-answer Internet products, users ask questions according to their own needs, and other users solve the problems and provide answers. The answers to the questions can be provided as search results to other users with similar questions, thus realizing knowledge sharing.

[0003] However, due to the openness of Internet products, the quality of answers varies greatly, so it is necessary to identify the quality of questions and answers. Summary of the invention

[0004] The present disclosure provides a method, apparatus, device and storage medium for determining question and answer quality.

[0005] According to one aspect of the present disclosure, a method for determining question and answer quality is provided, comprising:

[0006] Matching the question-answer pair to be identified with the candidate high-quality question-answer pair to obtain a target high-quality question-answer pair; wherein the question-answer pair to be identified includes a question to be identified and an answer to be identified, and the target high-quality question-answer pair includes a target high-quality question and a target high-quality answer;

[0007] The quality of the question-answer pair to be identified is determined based on a first correlation between the answer to be identified and the target high-quality answer, and a second correlation between the question to be identified and the target high-quality question.

[0008] According to one aspect of the present disclosure, a question and answer quality determination device is provided, comprising:

[0009] A target question-answer pair module is used to match the question-answer pair to be identified with the candidate high-quality question-answer pair to obtain a target high-quality question-answer pair; wherein the question-answer pair to be identified includes a question to be identified and an answer to be identified, and the target high-quality question-answer pair includes a target high-quality question and a target high-quality answer;

[0010] The quality determination module is used to determine the quality of the question-answer pair to be identified based on a first correlation between the answer to be identified and the target high-quality answer, and a second correlation between the question to be identified and the target high-quality question.

[0011] According to a third aspect, an electronic device is provided, the electronic device comprising:

[0012] at least one processor; and

[0013] a memory communicatively connected to the at least one processor; wherein,

[0014] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the question and answer quality determination method as described in any of the embodiments of the present application.

[0015] According to a fourth aspect, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute the question and answer quality determination method as described in any one of the embodiments of the present application.

[0016] The technology according to the present application can improve the accuracy of question and answer quality.

[0017] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present application.

[0019] Figure 1 is a flowchart of a method for determining question and answer quality provided according to an embodiment of the present application;

[0020] Figure 2 is a flowchart of another method for determining question and answer quality provided in an embodiment of the present application;

[0021] Figure 3 is a flowchart of another method for determining question and answer quality provided according to an embodiment of the present application;

[0022] Figure 4 is a structural schematic diagram of a question and answer quality determination device provided according to an embodiment of the present application;

[0023] Figure 5 It is a block diagram of an electronic device used to implement the question and answer quality determination method of an embodiment of the present application. DETAILED DESCRIPTION

[0024] The following is a description of exemplary embodiments of the present application in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0025] Figure 1 : is a flowchart of a method for determining the quality of a question and answer provided in an embodiment of the present application. This embodiment is applicable to situations where the quality of a question and answer needs to be evaluated. The method for determining the quality of a question and answer disclosed in this embodiment can be executed by an electronic device, specifically by a device for determining the quality of a question and answer, which can be implemented in software and / or hardware and configured in an electronic device. Figure 1 The question and answer quality determination method provided in this embodiment includes:

[0026] S110: Match the question-answer pair to be identified with the candidate high-quality question-answer pairs to obtain the target high-quality question-answer pair.

[0027] The question-answer pair to be identified includes the question to be identified and the answer to be identified, and the target high-quality question-answer pair includes the target high-quality question and the target high-quality answer. Correspondingly, the candidate high-quality question-answer pair includes the candidate high-quality question and the candidate high-quality answer.

[0028] The question and answer pairs to be identified can be questions and answers to the questions that need to determine the quality in question and answer products, for example, they can be question and answer pairs in question and answer communities, or they can be question and answer pairs on other forms of web pages on the Internet. In addition, question and answer pairs whose number of likes or shares exceeds a quantity threshold can be used as candidate high-quality question and answer pairs, and question and answer pairs provided by authoritative experts can also be used as candidate high-quality question and answer pairs. The embodiment of the present application does not specifically limit the source of the candidate high-quality question and answer pairs, and question and answer pairs that meet the high-quality conditions can be used as candidate high-quality question and answer pairs and added to the high-quality question and answer library. Moreover, there is no specific limitation on the high-quality conditions.

[0029] Specifically, the question to be identified can be matched with the candidate high-quality questions, and / or the answer to be identified can be matched with the candidate high-quality answers, and a target high-quality question-answer pair can be selected from the candidate high-quality question-answer pairs based on the matching results. Furthermore, the candidate high-quality question-answer pair with the highest matching degree can be used as the target high-quality question-answer pair. By introducing the target high-quality question-answer pair, it is convenient to use the target high-quality question-answer pair as a basis to determine the quality of the question-answer pair to be identified.

[0030] In the related art, the quality of the question-answer pair to be identified is determined directly based on the correlation between the answer to be identified and the question to be identified. However, this method has a limited scope of application and is only applicable to question-answer content recognition scenarios with low correlation. It cannot be applied to irrelevant answers caused by the correlation between the question and the answer, and has low accuracy in long answer scenarios. For example, the question is: Are apples delicious? The answer is: Apples are a kind of fruit. For another example, the question is how about Company A? The answer is: Introduction information of Company B, and Company A and Company B belong to the same industry field, and the introduction messages have a lot of overlapping content.

[0031] S120. Determine the quality of the question-answer pair to be identified based on a first correlation between the answer to be identified and the target high-quality answer, and a second correlation between the question to be identified and the target high-quality question.

[0032] The first correlation is used to represent the relationship between the answer to be identified and the target high-quality answer. The larger the first correlation is, the closer the two are. The second correlation is used to represent the relationship between the question to be identified and the target high-quality question. The larger the second correlation is, the closer the two are. Specifically, if both the first correlation and the second correlation are greater than the correlation threshold, the quality of the question-answer pair to be identified is qualified.

[0033] In an optional implementation, S120 includes: if the first correlation between the answer to be identified and the target high-quality answer is greater than a plagiarism similarity threshold, and the second correlation between the question to be identified and the target high-quality question is less than a semantic similarity threshold, then it is determined that the question-answer pair to be identified is irrelevant.

[0034] Among them, the plagiarism similarity threshold is used to determine whether the answer to be identified has the possibility of plagiarizing the target high-quality answer, and the semantic similarity threshold is used to determine whether the question to be identified and the target high-quality question belong to the same question. Both the plagiarism similarity threshold and the semantic similarity threshold can be empirical values.

[0035] Specifically, if the answer to be identified may plagiarize the target high-quality answer, and the question to be identified is not semantically related to the target high-quality question, then the question-answer pair to be identified is irrelevant. This technical feature can solve the irrelevant answer caused by copying similar answers. Such answers generally have a certain relevance, but it does not solve the problem, that is, it can improve the recognition accuracy of irrelevant answers, especially for the case where the answer to be identified is long. In addition, in order to balance the efficiency and accuracy of determining the quality of the answer, it can be distinguished according to the length of the answer to be identified. If the length is short, the quality of the question-answer pair to be identified can be directly determined based on the relevance between the answer to be identified and the question to be identified; if the length is long, the target high-quality question-answer pair can be introduced, and the quality of the question-answer pair to be identified can be determined based on the first relevance and the second relevance. In addition, a list of sensitive feature words can be set to determine whether the question to be identified and / or the answer to be identified includes sensitive feature words. If so, it is determined that the quality of the question-answer pair to be identified is low.

[0036] The technical solution of the embodiment of the present application can improve the accuracy of the quality of the question and answer pair to be identified, especially for scenarios with long answers, by introducing a target high-quality question and answer pair that matches the question and answer pair to be identified as a basis for determining quality, and combining a first correlation between the answer to be identified and the target high-quality answer, as well as a second correlation between the question to be identified and the target high-quality question to determine the quality of the question and answer pair to be identified.

[0037] Figure 2 is a flow chart of a method for determining the quality of a question and answer provided in an embodiment of the present application. This embodiment is an optional solution proposed on the basis of the above embodiment. Figure 2 The question and answer quality determination method provided in this embodiment includes:

[0038] S210: Determine a first similarity between the answer to be identified and a candidate high-quality answer in the candidate high-quality question and answer pair.

[0039] Among them, the cosine similarity between the text content of the answer to be identified and the text content of the candidate high-quality answer can be used as the first similarity between the two. Specifically, the text content of the answer to be identified and the text content of the candidate high-quality answer are segmented, the word vectors of the words are determined, and the word vectors are mean-optimized to obtain the vector representation of the answer to be identified and the vector representation of the candidate high-quality answer, and the cosine similarity between the two is determined.

[0040] S220. Determine a first correlation between the answer to be identified and the candidate high-quality answer based on the first similarity.

[0041] Specifically, the first similarity between the two can be directly used as the first correlation. Subsequently, by using the candidate high-quality answer that matches the answer to be identified as the target high-quality answer according to the first correlation, the similarity between the target high-quality answer and the answer to be identified can be improved, that is, the target high-quality answer that the answer to be identified may copy can be located.

[0042] In an optional implementation, S220 includes: determining a sentence repetition rate between the answer to be identified and the candidate high-quality answer; and determining a first correlation between the answer to be identified and the candidate high-quality answer based on the first similarity and the sentence repetition rate.

[0043] In this embodiment, the common sentences between the answer to be identified and the candidate high-quality answers can be determined, and the sentence repetition rate can be determined based on the number of common sentences. Specifically, the proportion of the number of common sentences in the answer to be identified and the candidate high-quality answers can be used as the sentence repetition rate of the answer to be identified and the sentence repetition rate of the candidate high-quality answers, respectively. Accordingly, the first correlation is determined based on the first similarity, the sentence repetition rate of the answer to be identified, and the sentence repetition rate of the candidate high-quality answers.

[0044] S230. Determine the target high-quality question-answer pair according to the first relevance.

[0045] Since the first correlation includes both the similarity between answers and the sentence repetition rate, the target high-quality answer that the answer to be identified may plagiarize can be located through the first correlation. The candidate high-quality question-answer pair to which the target high-quality answer belongs is used as the target high-quality question-answer pair.

[0046] S240: Determine the quality of the question-answer pair to be identified based on a first correlation between the answer to be identified and the target high-quality answer, and a second correlation between the question to be identified and the target high-quality question.

[0047] Specifically, if the first correlation is greater than the plagiarism similarity threshold, and the second correlation between the question to be identified and the target high-quality question is less than the semantic similarity threshold, it is determined that the question-answer pair to be identified is irrelevant.

[0048] The technical solution of the embodiment of the present application determines the first correlation between the answer to be identified and the candidate high-quality answers based on the first similarity and sentence repetition rate between the two, so as to measure whether the answer to be identified plagiarizes the candidate high-quality answers, and takes the candidate high-quality answer with the greatest possibility of plagiarism as the target high-quality answer, and obtains the target high-quality question associated with the target high-quality answer, which can provide the accuracy of the target high-quality answer, thereby improving the accuracy of the quality of the question and answer pairs to be identified.

[0049] Figure 3 is a flow chart of a method for determining the quality of a question and answer provided in an embodiment of the present application. This embodiment is an optional solution proposed on the basis of the above embodiment. Figure 3 The question and answer quality determination method provided in this embodiment includes:

[0050] S310: Match the question-answer pair to be identified with the candidate high-quality question-answer pairs to obtain a target high-quality question-answer pair.

[0051] Among them, the question-answer pair to be identified includes a question to be identified and an answer to be identified, and the target high-quality question-answer pair includes a target high-quality question and a target high-quality answer.

[0052] In an optional implementation, before matching the question-and-answer pair to be identified with the candidate high-quality question-and-answer pair, it also includes: determining the candidate high-quality question-and-answer pair based on the behavioral attribute data of the historical question-and-answer pairs; wherein the behavioral attribute data includes at least one of the following: the number of positive behaviors, producer information, time and site information.

[0053] Among them, positive behavior can be like behavior, comment behavior or sharing behavior; producer information can be the identity information of the provider of historical answers in the historical question and answer pairs, such as authoritative providers or non-authoritative providers; time refers to the time when the historical answer was provided. The earlier the time, the greater the probability that the historical answer is original, that is, the higher the quality of the historical answer; site information can be the site type of the question product to which the historical question and answer pair belongs, such as authoritative sites or non-authoritative sites. Based on the behavior attribute data, candidate high-quality question and answer pairs are mined from the historical question and answer pairs of each question and answer product to provide a basis for determining the quality of the question and answer pairs to be identified, so that the method for determining the quality of question and answer is highly universal.

[0054] S320: Determine a second similarity between the question to be identified and the target high-quality question.

[0055] Specifically, the cosine similarity between the text content of the question to be identified and the text content of the target high-quality question may be used as the second similarity between the two.

[0056] S330: Determine a second correlation between the question to be identified and the target high-quality question based on the second similarity.

[0057] Specifically, the second similarity between the two can be directly used as the second relevance. Subsequently, by determining whether the questions are similar based on the second relevance between the question to be identified and the target high-quality question, combined with the first relevance, irrelevant questions with similar answers but dissimilar questions can be identified.

[0058] In an optional implementation, S330 includes: determining the distance between the attribute information of the question to be identified and the attribute information of the target high-quality question; and determining the second correlation between the question to be identified and the target high-quality question based on the second similarity and the distance.

[0059] In an optional implementation manner, the attribute information includes at least one of the following: question type, question label, question keyword, and question industry field.

[0060] The question type can be digital, entity, time, method or definition, for example, when is the Chinese New Year a digital question, and what to eat during the Chinese New Year a entity question. Question tags can be the tags added by the questioner to the question, such as holidays or animals, etc. Question keywords can be obtained by performing named entity recognition on the question, and question industry fields can include economy, technology, games or automobiles, etc.

[0061] Furthermore, the priority of question type, question label, question keyword and question industry field decreases in turn, that is, the impact on the second relevance decreases in turn. Introducing attribute information in the process of determining the second relevance can improve the accuracy of the second relevance, that is, according to the second relevance, it can be determined whether the semantics of the to-be-identified question and the target high-quality question are similar.

[0062] S340: Determine the quality of the question-answer pair to be identified based on a first correlation between the answer to be identified and the target high-quality answer, and a second correlation between the question to be identified and the target high-quality question.

[0063] The technical solution of the embodiment of the present application, by introducing the attribute information of the question in the second relevance determination process, can determine whether the semantics between the question to be identified and the target high-quality question are similar according to the second relevance, thereby improving the accuracy of the quality of the question-answer pair to be identified. In addition, by determining the candidate high-quality question-answer pairs according to the behavioral attribute data of the historical question pairs, the versatility of the question-answer quality determination can be improved.

[0064] Figure 4 is a structural diagram of a device for determining the quality of a question and answer provided in an embodiment of the present application. Figure 4 The question and answer quality determination device 400 provided in the embodiment of the present application may include:

[0065] The target question-answer pair module 401 is used to match the question-answer pair to be identified with the candidate high-quality question-answer pair to obtain a target high-quality question-answer pair; wherein the question-answer pair to be identified includes a question to be identified and an answer to be identified, and the target high-quality question-answer pair includes a target high-quality question and a target high-quality answer;

[0066] The quality determination module 402 is used to determine the quality of the question-answer pair to be identified based on a first correlation between the answer to be identified and the target high-quality answer, and a second correlation between the question to be identified and the target high-quality question.

[0067] Optionally, the target question-answer pair module 401 includes:

[0068] A first similarity unit, used to determine a first similarity between the answer to be identified and a candidate high-quality answer in the candidate high-quality question-answer pair;

[0069] A first correlation unit, configured to determine a first correlation between the answer to be identified and the candidate high-quality answer according to the first similarity;

[0070] The target question-answer pair unit is used to determine the target high-quality question-answer pair according to the first relevance.

[0071] Optionally, the first relevance unit includes:

[0072] A repetition rate subunit, used to determine the sentence repetition rate between the answer to be identified and the candidate high-quality answer;

[0073] The first relevance subunit is used to determine the first relevance between the answer to be identified and the candidate high-quality answer according to the first similarity and the sentence repetition rate.

[0074] Optionally, the apparatus 400 further includes a second relevance module, and the second relevance module includes:

[0075] A second similarity unit, used to determine a second similarity between the to-be-identified question and the target high-quality question;

[0076] The second correlation unit is used to determine a second correlation between the to-be-identified question and the target high-quality question according to the second similarity.

[0077] Optionally, the second relevance unit includes:

[0078] A distance subunit, used to determine the distance between the attribute information of the problem to be identified and the attribute information of the target high-quality problem;

[0079] The second relevance subunit is used to determine a second relevance between the to-be-identified question and the target high-quality question according to the second similarity and the distance.

[0080] Optionally, the attribute information includes at least one of the following: question type, question label, question keyword and question industry field.

[0081] Optionally, the device 400 further includes:

[0082] The candidate high-quality module is used to determine the candidate high-quality question-answer pairs based on the behavioral attribute data of the historical question-answer pairs; wherein the behavioral attribute data includes at least one of the following: the number of positive behaviors, producer information, time and site information.

[0083] Optionally, the quality determination module 402 is specifically configured to:

[0084] If the first correlation between the answer to be identified and the target high-quality answer is greater than the plagiarism similarity threshold, and the second correlation between the question to be identified and the target high-quality question is less than the semantic similarity threshold, it is determined that the question-answer pair to be identified is irrelevant.

[0085] The technical solution of the embodiment of the present application can improve the accuracy of the quality of question and answer by identifying irrelevant answers that have certain relevance but no characteristic word hits by using the fingerprint information of the whole network of question and answer (i.e., high-quality question and answer information) to identify relevant irrelevant answers, and taking similar answers but dissimilar questions as an indicator to measure irrelevant answers, and can be applied to the identification of low-quality irrelevant answers that have certain relevance but no characteristic word hits.

[0086] According to an embodiment of the present application, the present application also provides an electronic device and a readable storage medium.

[0087] like Figure 5 , is a block diagram of an electronic device according to a method for determining the quality of questions and answers according to an embodiment of the present application. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0088] like Figure 5 As shown, the electronic device includes: one or more processors 501, a memory 502, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are interconnected using different buses and can be installed on a common mainboard or installed in other ways as needed. The processor can process instructions executed in the electronic device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In other embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple electronic devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 5 A processor 501 is taken as an example.

[0089] The memory 502 is a non-transitory computer-readable storage medium provided in the present application. The memory stores instructions executable by at least one processor to enable the at least one processor to perform the method for determining the quality of question and answer provided in the present application. The non-transitory computer-readable storage medium of the present application stores computer instructions, which are used to enable a computer to perform the method for determining the quality of question and answer provided in the present application.

[0090] The memory 502 is a non-transient computer-readable storage medium that can be used to store non-transient software programs, non-transient computer executable programs and modules, such as the program instructions / modules corresponding to the method for determining the quality of question and answer in the embodiment of the present application (for example, the attached Figure 4 The processor 501 executes various functional applications of the server and question and answer quality determination by running the non-transient software programs, instructions and modules stored in the memory 502, that is, the question and answer quality determination method in the above method embodiment is implemented.

[0091] The memory 502 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and applications required for at least one function; the data storage area may store data created based on the use of the electronic device determined according to the quality of the question and answer, etc. In addition, the memory 502 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory 502 may optionally include a memory remotely disposed relative to the processor 501, and these remote memories may be connected to the electronic device determined by the quality of the question and answer via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0092] The electronic device of the question-answer quality determination method may further include: an input device 503 and an output device 504. The processor 501, the memory 502, the input device 503 and the output device 504 may be connected via a bus or other means. Figure 5 The example of connecting through bus is taken in the following.

[0093] The input device 503 can receive input digital or character information, and generate key signal input related to user settings and function control of the electronic device for question and answer quality determination, such as a touch screen, a keypad, a mouse, a track pad, a touch pad, an indicator bar, one or more mouse buttons, a trackball, a joystick and other input devices. The output device 504 may include a display device, an auxiliary lighting device (e.g., an LED) and a tactile feedback device (e.g., a vibration motor), etc. The display device may include, but is not limited to, a liquid crystal display (LCD), a light emitting diode (LED) display and a plasma display. In some embodiments, the display device may be a touch screen.

[0094] Various implementations of the systems and techniques described herein can be realized in digital electronic circuit systems, integrated circuit systems, dedicated ASICs (application specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0095] These computer programs (also referred to as programs, software, software applications, or code) include machine instructions for programmable processors and can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or means (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0096] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0097] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0098] A computer system may include clients and servers. Clients and servers are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship to each other.

[0099] The technical solution of the embodiment of the present application introduces a target high-quality question-answer pair that matches the question-answer pair to be identified as the basis for quality identification, identifies relevant irrelevant answers by using the whole network question-answer fingerprint information (i.e., high-quality question-answer information), and uses similar answers but dissimilar questions as an indicator to measure irrelevant answers, thereby identifying irrelevant answers and improving the accuracy of question-answer quality. It can be applied to the identification of low-quality irrelevant answers that have certain relevance but no feature word hits.

[0100] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for determining the quality of a question and answer, comprising: Matching the question-answer pair to be identified with the candidate high-quality question-answer pair to obtain a target high-quality question-answer pair; wherein the question-answer pair to be identified includes a question to be identified and an answer to be identified, and the target high-quality question-answer pair includes a target high-quality question and a target high-quality answer; If the first correlation between the answer to be identified and the target high-quality answer is greater than the plagiarism similarity threshold, and the second correlation between the question to be identified and the target high-quality question is less than the semantic similarity threshold, then it is determined that the question-answer pair to be identified is a non-answer to the question; wherein the first correlation is used to represent the relationship between the answer to be identified and the target high-quality answer; the plagiarism similarity threshold is used to determine whether the answer to be identified has the possibility of plagiarizing the target high-quality answer; the second correlation is used to represent the relationship between the question to be identified and the target high-quality question; the semantic similarity threshold is used to determine whether the question to be identified and the target high-quality question are the same question.

2. The method according to claim 1, wherein: The step of matching the question-answer pair to be identified with the candidate high-quality question-answer pair to obtain the target high-quality question-answer pair includes: Determining a first similarity between the answer to be identified and a candidate high-quality answer in the candidate high-quality question-answer pair; Determining a first correlation between the answer to be identified and the candidate high-quality answer according to the first similarity; The target high-quality question-answer pair is determined according to the first relevance.

3. The method according to claim 2, wherein: The determining, based on the first similarity, a first correlation between the answer to be identified and the candidate high-quality answer includes: Determine the sentence repetition rate between the answer to be identified and the candidate high-quality answer; A first correlation between the answer to be identified and the candidate high-quality answer is determined according to the first similarity and the sentence repetition rate.

4. The method according to claim 1, further comprising: Determining a second similarity between the to-be-identified question and the target high-quality question; According to the second similarity, a second correlation between the to-be-identified question and the target high-quality question is determined.

5. The method according to claim 4, wherein: Determining a second correlation between the to-be-identified question and the target high-quality question according to the second similarity includes: Determine the distance between the attribute information of the problem to be identified and the attribute information of the target high-quality problem; A second correlation between the to-be-identified question and the target high-quality question is determined according to the second similarity and the distance.

6. The method according to claim 5, wherein: The attribute information includes at least one of the following: question type, question label, question keyword and question industry field.

7. The method according to any one of claims 1 to 6, before matching the question-answer pair to be identified with the candidate high-quality question-answer pair, further comprising: The candidate high-quality question-answer pairs are determined based on the behavioral attribute data of the historical question-answer pairs; wherein the behavioral attribute data includes at least one of the following: the number of positive behaviors, producer information, time and site information.

8. A device for determining the quality of a question and answer, comprising: A target question-answer pair module is used to match the question-answer pair to be identified with the candidate high-quality question-answer pair to obtain a target high-quality question-answer pair; wherein the first correlation is used to represent the relationship between the answer to be identified and the target high-quality answer; the question-answer pair to be identified includes the question to be identified and the answer to be identified; the second correlation is used to represent the relationship between the question to be identified and the target high-quality question; the target high-quality question-answer pair includes the target high-quality question and the target high-quality answer; A quality determination module, configured to determine the quality of the to-be-identified question-answer pair according to a first correlation between the to-be-identified answer and the target high-quality answer, and a second correlation between the to-be-identified question and the target high-quality question; Wherein, the quality determination module is specifically used for: If the first correlation between the answer to be identified and the target high-quality answer is greater than the plagiarism similarity threshold, and the second correlation between the question to be identified and the target high-quality question is less than the semantic similarity threshold, then it is determined that the question-answer pair to be identified is irrelevant. The plagiarism similarity threshold is used to determine whether the answer to be identified has the possibility of plagiarizing the target high-quality answer, and the semantic similarity threshold is used to determine whether the question to be identified and the target high-quality question are the same question.

9. The device according to claim 8, wherein: The target question-answer pair module includes: A first similarity unit, used to determine a first similarity between the answer to be identified and a candidate high-quality answer in the candidate high-quality question-answer pair; A first correlation unit, configured to determine a first correlation between the answer to be identified and the candidate high-quality answer according to the first similarity; The target question-answer pair unit is used to determine the target high-quality question-answer pair according to the first relevance.

10. The device according to claim 9, wherein: The first relevance unit comprises: A repetition rate subunit, used to determine the sentence repetition rate between the answer to be identified and the candidate high-quality answer; The first relevance subunit is used to determine the first relevance between the answer to be identified and the candidate high-quality answer according to the first similarity and the sentence repetition rate.

11. The apparatus according to claim 8, further comprising a second relevance module, wherein the second relevance module comprises: A second similarity unit, used to determine a second similarity between the to-be-identified question and the target high-quality question; The second correlation unit is used to determine a second correlation between the to-be-identified question and the target high-quality question according to the second similarity.

12. The device according to claim 11, wherein The second relevance unit comprises: A distance subunit, used to determine the distance between the attribute information of the problem to be identified and the attribute information of the target high-quality problem; The second relevance subunit is used to determine a second relevance between the to-be-identified question and the target high-quality question according to the second similarity and the distance.

13. The device according to claim 12, wherein: The attribute information includes at least one of the following: question type, question label, question keyword and question industry field.

14. The device according to any one of claims 8 to 13, further comprising: The candidate high-quality module is used to determine the candidate high-quality question-answer pairs based on the behavioral attribute data of the historical question-answer pairs; wherein the behavioral attribute data includes at least one of the following: the number of positive behaviors, producer information, time and site information.

15. An electronic device, wherein: include: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.

16. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 7.

Images