CN110765247A

CN110765247A - Input prompting method and device for question-answering robot

Info

Publication number: CN110765247A
Application number: CN201910940923.8A
Authority: CN
Inventors: 杨明晖; 陈晓军
Original assignee: Alipay Hangzhou Information Technology Co Ltd
Current assignee: Alipay Hangzhou Information Technology Co Ltd
Priority date: 2019-09-30
Filing date: 2019-09-30
Publication date: 2020-02-07
Anticipated expiration: 2039-09-30
Also published as: CN110765247B

Abstract

The specification provides an input prompting method and device for a question and answer robot, wherein the method comprises the following steps: the method comprises the steps of carrying out word segmentation and normalization preprocessing on question input information received by a question answering robot, selecting candidate prompt information from a knowledge point database based on the preprocessed information, screening target prompt information by taking text similarity and frequency information of the candidate question information, namely flow ratio, as a measurement index, and recommending the target prompt information to a user.

Description

Input prompting method and device for question-answering robot

Technical Field

The specification belongs to the technical field of computers, and particularly relates to an input prompting method and device for a question and answer robot.

Background

With the development of computer technology, more and more intelligent robots appear in life and work of people, wherein a question-answering robot is a common robot. The question and answer robot can be used in customer service business, and is generally good at solving simple and standard questions and not good at processing complex and fuzzy questions. Since some users are unfamiliar with the characteristics of the robot, complex problems or fuzzy problems are often transmitted, which increases the processing difficulty of the robot and reduces the user experience.

Disclosure of Invention

An object of an embodiment of the present specification is to provide an input prompt method and apparatus for a question and answer robot, and a question and answer robot, which implement accurate recommendation of input prompt information for a user, and reduce workload of the user and data processing amount of the question and answer robot.

In one aspect, an embodiment of the present specification provides an input prompt method for a question and answer robot, including:

performing word segmentation and normalization processing on the received question input information to obtain preprocessed question information;

acquiring a target knowledge point title matched with the preprocessed question information from a knowledge point database, and taking the acquired target knowledge point title as candidate prompt information;

determining the similarity between the candidate prompt information and the pre-processing question extracting information and the flow rate ratio of the candidate prompt information in a specified time range;

and screening target prompt information from the candidate prompt information according to the similarity and the traffic ratio corresponding to the candidate prompt information, and recommending the target prompt information to a user.

In another aspect, the present specification provides an input presentation apparatus for a question-answering robot, including:

the preprocessing module is used for carrying out word segmentation and normalization processing on the received question input information to obtain preprocessed question information;

the candidate information acquisition module is used for acquiring a target knowledge point title matched with the preprocessed question information from a knowledge point database, and taking the acquired target knowledge point title as candidate prompt information;

the similarity frequency determining module is used for determining the similarity between the candidate prompt information and the pre-processing question extracting information and the traffic ratio of the candidate prompt information in a specified time range;

and the prompt information screening module is used for screening target prompt information from the candidate prompt information according to the similarity and the traffic ratio corresponding to the candidate prompt information and recommending the target prompt information to a user.

In yet another aspect, the present specification provides an input prompting device for a question and answer robot, comprising: the input prompting method for the question answering robot comprises at least one processor and a memory for storing processor executable instructions, wherein the processor executes the instructions to realize the input prompting method for the question answering robot.

In another aspect, an embodiment of the present specification provides a question-answering robot, including: the input prompting method for the question-answering robot is realized when the processor executes the instruction, and corresponding answers are output according to clicking operation of a user on target prompting information.

The input prompting method, the input prompting device, the processing equipment and the question-answering robot for the question-answering robot, which are provided by the specification, preprocess the question input information received by the question-answering robot, select candidate prompting information to be searched in a knowledge point database based on the preprocessed information, and screen out target prompting information by taking text similarity and frequency information of the candidate prompting information, namely flow ratio, as a measurement index, so that the accuracy of screening the target prompting information is improved, and the target prompting information is closer to the requirements of users. And recommending the obtained target prompt information to the user for selection by the user, so that the workload of the user is reduced. Meanwhile, the recommended target prompt information generally conforms to the question format of the question-answering robot, so that the data processing amount of the question-answering robot is reduced, and the capability of the question-answering robot for accurately answering the user questions is improved.

Drawings

In order to more clearly illustrate the embodiments of the present specification or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only some embodiments described in the present specification, and for those skilled in the art, other drawings can be obtained according to the drawings without any creative effort.

FIG. 1 is a schematic flow chart diagram illustrating an input prompt method for a question and answer robot in one embodiment of the present disclosure;

FIG. 2 is a schematic flow chart diagram of a question-answering robot input prompt method in yet another embodiment of the present specification;

FIG. 3 is a block diagram of an embodiment of an input prompt apparatus for a question-answering robot according to the present disclosure;

FIG. 4 is a schematic structural diagram of an input prompting device for a question-answering robot in another embodiment of the present specification;

FIG. 5 is a schematic structural diagram of an input prompting device for a question-answering robot in yet another embodiment of the present specification;

fig. 6 is a block diagram of a hardware configuration of an input prompt server for the question and answer robot in one embodiment of the present specification.

Detailed Description

In order to make those skilled in the art better understand the technical solutions in the present specification, the technical solutions in the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is obvious that the described embodiments are only a part of the embodiments of the present specification, and not all of the embodiments. All other embodiments obtained by a person skilled in the art based on the embodiments in the present specification without any inventive step should fall within the scope of protection of the present specification.

The question-answering robot can be used for answering simple text questions, and is applied to a plurality of service industries, such as: the customer service of the shopping website usually has a question and answer robot, and some common and simple questions of the user can be answered by the question and answer robot, so that the amount of manual labor is reduced.

The embodiment of the specification provides an input prompting method for a question and answer robot, which can recommend question information for a user according to the question input information of the user, and can be understood as progressive prompting, namely, the system can recommend content which is possibly input for the user near an input box in the input process of the user, so that the input of the user is facilitated. In some search engines, related search requests are recommended when the user inputs the search requests, so that the user input is assisted, and the recommended content is derived from historical search requests. However, the search engine calls are extremely large, and several engines call hundreds of millions or more times per day, so there are enough historical requests for recommendations. However, for the question and answer robot, on one hand, the user request amount is small, and all the input of the user is difficult to cover; on the other hand, most texts searched by a user in a search engine have results, but only part of questions in the question-answering robot can find answers, so that the progressive prompt is to avoid recommending questions that the question-answering robot cannot answer, and therefore, the progressive prompt scheme of the search engine is not suitable for the question-answering robot.

The input prompting method for the question and answer robot in the specification can be applied to a client or a server, and the client can be an electronic device such as a smart phone, a tablet computer, a smart wearable device (a smart watch, virtual reality glasses, a virtual reality helmet and the like), a smart vehicle-mounted device and the like.

Fig. 1 is a schematic flowchart of an input prompting method for a question and answer robot in an embodiment of the present specification, and as shown in fig. 1, the input prompting method for the question and answer robot in an embodiment of the present specification may include:

and 102, performing word segmentation and normalization processing on the received question input information to obtain preprocessed question information.

In a specific implementation process, the question input information may represent text information input by the user in the question and answer robot, and may also be voice information. The question input information may be an incomplete question, and the embodiment of the present specification may acquire the question input information input by the user in real time in the process of user input, for example: when the user inputs 'flower value how … …' is not input yet, the corresponding question input information can be obtained, so that the question which the user wants to input is recommended to the user in time according to the content input by the user, the user can ask the question conveniently, and the workload is reduced.

After the question input information is obtained, natural language processing of word segmentation and normalization can be carried out on the question input information, and preprocessed question information is obtained. Word segmentation may be understood as dividing a sentence into a plurality of words according to the structure or part of speech of the sentence. Such as: the 'flower shape' is divided into two words of 'flower' and 'how'. Normalization may be understood as converting the questioning input information into text in a specified format, such as: punctuation marks in the sentence can be deleted, all English letters are unified into lower case, and the like, so that the subsequent processing is facilitated. Taking the information after word segmentation and normalization as the pre-processed question information, such as: the 'flower how' is obtained after the word segmentation and normalization processing.

And 104, acquiring a target knowledge point title matched with the preprocessed question information from a knowledge point database, and taking the acquired target knowledge point title as candidate prompt information.

In a particular implementation, a knowledge point typically includes a title and a body that describe a business function or rule; in the question-answering robot, the 'question' is commonly used as a title, the 'answer' is used as a text, the text of the knowledge point can be replied to the user by finding out the title of the knowledge point which is most similar to the question of the user, and the knowledge point database can represent the collection of the knowledge points. In some embodiments of the present specification, a knowledge point database may be configured for the question and answer robot in advance, and the knowledge point database may include a knowledge point title and a corresponding answer. The content in the knowledge point database can be determined according to the application scenario of the question-answering robot, such as: the knowledge point database may be constructed based on common knowledge of services, frequently-used question and answer data obtained by web crawlers, and the like, and the specific construction method is not specifically limited in the embodiments of this specification.

After the received input question information is subjected to word segmentation and normalization processing, a target knowledge point title matched with the preprocessed question information obtained after processing can be obtained from the knowledge point database, and the obtained target knowledge point title is used as candidate prompt information. Here, "match" can be understood as whether the knowledge point header carries the pre-processing question information, such as: the pre-processing question information in the above embodiment is "how you spend", a knowledge point title with "how you spend" may be obtained from the knowledge point database, for example: how to repay the flower, how to open the flower, etc. as the target knowledge point title.

And 106, determining the similarity between the candidate prompt information and the pre-processing question-raising information and the flow rate ratio of the candidate prompt information in a specified time range.

In a specific implementation process, the text similarity can be understood as the similarity between two texts, and the higher the similarity is, the more similar the meanings of the two texts are. The similarity between the candidate prompt information and the pre-processing problem-raising information may be determined by using a text similarity algorithm, where the text similarity algorithm may be selected according to an actual use situation, and in some embodiments of the present specification, a bert (bidirectional encoder responses from transform) model, or a BM25 (generally used for search relevance halving), a WMD (word mover's distance, which is a method for calculating a distance between sentences), an ESIM (Enhanced LSTM for Natural language inference, a Long Short-Term Memory network (LSTM)) model, or the like may be used to perform similarity calculation on the candidate prompt information and the pre-processing problem-raising information, so as to determine the similarity between the candidate prompt information and the pre-processing problem-raising information.

After the candidate prompt information is obtained, the traffic ratio of each candidate prompt information can be counted, the traffic ratio can be understood as frequency information or frequency information corresponding to each candidate prompt information, and the condition that each candidate prompt information is used in the whole knowledge point database can be reflected. The traffic ratio of the candidate prompt message may be calculated according to the condition that each knowledge point in the knowledge point database is used, for example: the traffic ratio of the candidate presentation information a is the number of times of use of the candidate presentation information a/the number of times (or number) of all the knowledge point titles used by the user.

And 108, screening target prompt information from the candidate prompt information according to the similarity and the traffic ratio corresponding to the candidate prompt information, and recommending the target prompt information to a user.

After the similarity between each candidate prompt message and the pre-processing question-raising message and the traffic ratio of each candidate prompt message are determined, each candidate prompt message can be comprehensively scored based on the similarity and the traffic ratio, and one or more target prompt messages can be screened out. Such as: the similarity and traffic to information ratio may be converted into corresponding scores, such as: determining a function for converting the similarity and the traffic information into scores based on historical data, converting the similarity and the traffic information into corresponding scores based on the function, determining a comprehensive score corresponding to each candidate prompt information according to the converted scores, and considering that the candidate prompt information is closer to the question information desired by the user when the comprehensive score is higher. Feeding back the screened target prompt information to the user as follows: and displaying the target prompt information in the input box accessory for the user to select and click.

The input prompting method for the question-answering robot provided by the embodiment of the specification preprocesses the question input information received by the question-answering robot, selects and searches candidate prompting information in the knowledge point database based on the preprocessed information, and screens out target prompting information by taking text similarity and frequency information of the candidate prompting information, namely flow ratio, as a measurement index, so that the accuracy of screening the target prompting information is improved, and the target prompting information is closer to the requirements of users. And recommending the obtained target prompt information to the user for selection by the user, so that the workload of the user is reduced. Meanwhile, the recommended target prompt information generally conforms to the question format of the question-answering robot, so that the data processing amount of the question-answering robot is reduced, and the probability of accurately answering the user questions by the question-answering robot is improved.

On the basis of the foregoing embodiments, in some embodiments of the present specification, the method may further include:

and acquiring a target high-frequency question matched with the preprocessed question information from a high-frequency question library, and taking the acquired target high-frequency question and the target knowledge point title as candidate prompt information.

In a specific implementation process, a high-frequency question bank may be constructed in advance according to the question information of the user, and the high-frequency question bank may include questions with a relatively high question-asking frequency of the user and may also include answers corresponding to the questions. After the question input information input by the user is received and processed, the target knowledge point title and the target high-frequency question matched with the preprocessed question information can be acquired from the knowledge point database and the high-frequency question database respectively according to the preprocessed question information obtained after processing. Namely, searching the knowledge point title and the high-frequency question with the preprocessed question information from the knowledge point database and the high-frequency question database respectively as candidate prompt information. Specifically, an ES (elastic search, a search server) can be used to search for a knowledge point title and a high-frequency problem.

The embodiment of the specification can not only search candidate prompt information from the knowledge point database, but also search high-frequency questions matched with input question information from the high-frequency question database to serve as the candidate prompt information, so that the content of the prompt information is expanded, and the coverage rate and the accuracy of the recommended prompt information are improved.

In some embodiments of the present disclosure, the high frequency question bank may be constructed by the following method:

recording question information and answer information received by the question and answer robot,

according to the question information and the answering information, counting the question information successfully answered by the question and answer robot;

and taking the question information with the frequency greater than a preset threshold value or with the frequency ranked in a specified name in the question information which is successfully answered as a high-frequency question, and constructing the high-frequency question library.

In a specific implementation process, when the question-answering robot answers the question of the user, the question-answering robot may record the received question information and the corresponding answer information, where the answer information may include an answer given by the question-answering robot and feedback information of the user, and the feedback information may include information about whether a subsequent question of the user is related to a current question and whether the answer is directly fed back by the user. Some users may present complicated questions or may not be within the service range of the question-answering robot, and the question-answering robot may not answer or answer incorrectly. Some embodiments of the present specification may calculate, according to the question information and the answer information received by the question-answering robot, the question information that the question-answering robot can answer, that is, the question information that the answer is successful. And taking the question information with the frequency greater than a preset threshold value or the frequency ranking within a specified name range in the question information which is successfully answered as a high-frequency question, and taking the set of all the high-frequency questions as a high-frequency question library.

It should be noted that, according to the received question information and the corresponding answer information that are continuously recorded by the question answering robot, the high-frequency question bank may be continuously updated, and an update period of the high-frequency question bank may be set or the high-frequency question bank may be updated in real time.

In the embodiment of the specification, the high-frequency question bank is statistically constructed based on the question information and the corresponding answer information received by the question and answer robot, the question information in the dynamically maintained high-frequency question bank can better reflect the condition of the question of the user, and the coverage rate and the accuracy of the recommended prompt information are improved.

On the basis of the above embodiments, in some embodiments of the present specification, the performing word segmentation and normalization processing on the received question input information to obtain preprocessed question information includes:

performing word segmentation and normalization processing on the received question input information to obtain input text processing information;

synonym expansion is carried out on the input text processing information by utilizing a synonym table, and synonym information of the input text processing information is obtained;

and taking the synonymous information and the input text processing information as the pre-processing question information.

In a specific implementation process, when preprocessing the received input question information, the received question input information may be subjected to word segmentation and normalization processing to obtain input text processing information, where the process of word segmentation and normalization processing refers to the description of the above embodiment and is not described herein again. And performing synonymy expansion on the obtained input text processing information by using the synonym table, namely acquiring synonyms of the input text processing information from the synonym table and determining the synonym of the input text processing information. The synonym table may include correspondence between synonyms, and synonyms may be understood as words having the same or similar meanings. For example: "how" and "how" can be understood as synonyms, and after synonymy expansion of "how flower" in the above embodiment, synonymy information of "how flower" can be obtained. And the input text processing information obtained after word segmentation and preprocessing and the synonymy information obtained after synonymy expansion are used as the preprocessed question information. Such as: the 'how much of flower' and 'how much of flower' in the above embodiment are used as the pre-processing question information, and the subsequent candidate question information and the target question information are screened.

According to the implementation of the description, the question information input by the user is synonymously expanded, the secondary high rate of the candidate prompt information is increased, and the accuracy of target prompt information recommendation is improved.

On the basis of the foregoing embodiments, in some embodiments of this specification, the screening target prompt information from the candidate prompt information according to the similarity and the traffic ratio corresponding to the candidate prompt information may include:

respectively carrying out barrel separation processing on the similarity and the traffic ratio, and respectively converting the similarity and the traffic ratio into feature vectors with specified dimensions;

splicing the feature vectors of similarity and flow ratio corresponding to the same candidate prompt information to obtain an input feature vector;

scoring each candidate prompt message based on the input feature vector by using a prompt message scoring model;

and sorting candidate prompt information with the score larger than a preset threshold or the score from high to low in a preset ranking as the target prompt information.

In a specific implementation, the bucket division processing can be understood as a vector conversion technology, which can map one-dimensional data to a vector of a specified dimension. In an embodiment of the present description, the similarity and the traffic ratio corresponding to each candidate prompt information may be respectively subjected to bucket division, and the similarity and the traffic ratio are respectively converted into feature vectors of specified dimensions. And splicing the feature vector of the similarity conversion and the feature vector of the flow ratio conversion of the same candidate prompt information to obtain an input feature vector corresponding to the candidate prompt information. For example: if the similarity s and the flow ratio r of the candidate prompt information A are subjected to barrel dividing operation, a 5-dimensional feature vector corresponding to the similarity s and a 5-dimensional feature vector corresponding to the flow ratio r are obtained, the two 5-dimensional feature vectors are spliced to obtain a 10-dimensional vector, and the 10-dimensional vector can be used as an input feature vector of the candidate prompt information A.

The prompt information scoring model may be configured to score input feature vectors corresponding to the candidate prompt information, and after the input feature vectors corresponding to the candidate prompt information are input to the prompt information scoring model, scores corresponding to the candidate prompt information may be obtained, where the scores may be greater than a specified threshold, or the scores may be sorted from high to low, and the candidate prompt information ranked in a preset ranking is used as the target prompt information. The prompt information scoring model can be obtained by performing model training and constructing according to historical question information of a user and by combining manual marking information and the like. The hint information scoring model may employ a Logistic Regression model (LR) or a tree model such as: GBDT (Gradient boosting decision Tree), depth models such as DNN (Convolutional Neural Network model), and other models may also be used, and the embodiments of the present specification are not particularly limited.

In the embodiment of the description, candidate prompt information is converted into multi-dimensional feature vectors through barrel dividing operation, and then the candidate prompt information is graded and screened by utilizing a prompt information grading model to obtain target prompt information meeting conditions and recommended to a user, so that automatic prompt of input information of a question and answer robot is realized.

In some embodiments of the present specification, the prompt information scoring model may be further optimized according to click data of the target prompt information. Namely, the click data of the target prompt information is used as sample data, and the prompt information scoring model is optimally trained, so that the output accuracy of the prompt information scoring model is improved, and the accuracy of the prompt information is further improved.

Fig. 2 is a schematic flow chart of a question-answering robot input prompting method in another embodiment of this specification, and a method for determining the prompting information in this specification is specifically described below with reference to fig. 2:

when the user inputs the question information in the input box of the question-answering robot, the progressive prompt triggers the call. And sequentially executing preprocessing, recalling and reordering, and finally outputting a plurality of results to recommend to a user. In the following, a specific method of the embodiment of the present specification will be described by taking the question input information of "how to flower" as an example:

firstly, preprocessing: and in the preprocessing link, performing word segmentation, normalization and synonymy expansion on the question input information input by the user. The word segmentation means to segment the input sentence into a plurality of words, namely to divide 'flower how' into 'flower' and 'how' into two words. In the normalization process, punctuations in sentences are deleted, all English letters are unified into lowercase, and in one embodiment of the specification, "what is spent" is still the normalization process without other special treatment. Synonymy expansion, i.e., expansion of an input sentence into a synonymy sentence, can be implemented by a synonym table. Such as: "how" has the synonym "how", so after synonymy expansion, two phrases of "how flower" and "how flower" are output. The purpose of synonymy expansion is to increase the candidates obtained in the recall stage and improve the coverage rate.

II, recalling: the recall phase uses the inverted index ElasticSearch, and two types of data are contained in the ES index:

1. knowledge point title: when the user configures the question-answering robot, some knowledge points can be configured for the robot (namely, a knowledge point database is constructed) for answering the user question. Simple knowledge points are similar to FAQ (Frequently asked questions), containing a title and an answer to the title, e.g., "how many customer service calls? "-" 95188 ", the former is the title of the knowledge point and will be recommended to the user who entered the relevant text.

2. High frequency problem: when the question answering robot answers the question of the user, the question answering robot records the question information of the user. The system will count the questions that the robot has the ability to solve, where high frequency questions will be recommended to the user who entered the relevant text.

The user inputs 'how of flower', obtains 'how of flower' and 'how of flower' after preprocessing, and recalls knowledge point titles or high-frequency problems such as 'how of flower', 'how of flower repayment' and 'how of flower' from the knowledge point database and the high-frequency problem database through the ES.

Thirdly, reordering: step two recalls a plurality of knowledge point titles or high-frequency questions related to the user input, then sorts the knowledge point titles or high-frequency questions, and selects Top K (Top K) candidates (candidate prompt information, the knowledge point titles or high-frequency questions obtained in step two) to output. The ranking will use two types of features, text similarity and candidate frequency, respectively. The text similarity model requires to input a sentence q and a candidate c, the model records the similarity degree of the calculated q and c as s, and the similarity degree is a score in a [0,1] interval. The candidate frequency is the frequency at which the candidate is true on the line. The candidate item contains two parts, namely a knowledge point title and a user high-frequency problem. And (5) counting the flow rate ratio of the knowledge point title and the high-frequency problem in the last week, and recording as r. Similarly, r is also in the interval [0,1 ]. And finally, performing barrel dividing operation on the text similarity s and the candidate frequency r to respectively obtain two 5-dimensional characteristic bin _ s and bin _ r. In some embodiments of the present description, the bucket partitioning scheme may be: the | s range | bin _ s takes the value | as: | [0,0.2) | [ 10000 ] |, | [0.2,0.4) | [ 01000 ] |, | [0.4,0.6) | [ 00100 ] |, | [0.6,0.8) | [ 00010 ] |, | [0.8,1] | [ 00001 ] |. The range | bin _ r takes | such as: | [0,0.0001) | [ 10000 ] |, | [0.0001,0.001) | [ 01000 ] |, | [0.001,0.01) | [ 00100 ] |, | [0.01,0.1) | [ 00010 ] |, | [0.1,1] | [ 00001 ] |. And (3) splicing bin _ s and bin _ r to obtain 10-dimensional characteristics, training an LR score calculation model for calculating scores, and outputting TOP K results with the highest scores as recommended target prompt information. And feeding back the obtained TOP K target prompt information to the user, and displaying the TOP K target prompt information near an input box of the question and answer robot for the user to select.

The BERT model can be based on a basic structure of the Chinese BERT model, and manual marking data are utilized for fine adjustment to obtain a text similarity model. The LR scoring model can be trained on manual marking data and continuously optimized through on-line actual click data.

In the method provided by the embodiment of the specification, in the recall stage, besides relying on the knowledge base maintained manually, a high-frequency problem base can be maintained through the system and used for expanding recommendable contents, so that the coverage rate is greatly improved. In the sorting stage, the text similarity characteristics are used, and the frequency information of the candidate items is used, so that the problems that the knowledge point titles with shorter length have higher similarity when the user inputs the short information, and the knowledge point titles with high frequency and longer length are difficult to recommend to the user are solved. And parameters are adjusted through the click log, the most helpful recommendation is provided for the user, and the click rate is improved.

In the present specification, each embodiment of the method is described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. The relevant points can be obtained by referring to the partial description of the method embodiment.

Based on the input prompting method for the question and answer robot, one or more embodiments of the specification further provide an input prompting device for the question and answer robot. The apparatus may include systems (including distributed systems), software (applications), modules, components, servers, clients, etc. that use the methods described in the embodiments of the present specification in conjunction with any necessary apparatus to implement the hardware. Based on the same innovative conception, embodiments of the present specification provide an apparatus as described in the following embodiments. Since the implementation scheme of the apparatus for solving the problem is similar to that of the method, the specific apparatus implementation in the embodiment of the present specification may refer to the implementation of the foregoing method, and repeated details are not repeated. As used hereinafter, the term "unit" or "module" may be a combination of software and/or hardware that implements a predetermined function. Although the means described in the embodiments below are preferably implemented in software, an implementation in hardware, or a combination of software and hardware is also possible and contemplated.

Specifically, fig. 3 is a schematic block structural diagram of an embodiment of the input prompting device for a question-answering robot provided in this specification, and as shown in fig. 3, the input prompting device for a question-answering robot provided in this specification may include: the device comprises a preprocessing module 31, a candidate information acquisition module 32, a similarity frequency determination module 33 and a prompt information screening module 34, wherein:

the preprocessing module 31 may be configured to perform word segmentation and normalization processing on the received question input information to obtain preprocessed question information;

the candidate information obtaining module 32 may be configured to obtain a target knowledge point title matched with the preprocessed question information from the knowledge point database, and use the obtained target knowledge point title as candidate prompt information;

a similarity frequency determining module 33, configured to determine a similarity between the candidate prompt information and the pre-processing question extracting information, and a traffic ratio of the candidate prompt information in a specified time range;

the prompt information screening module 34 may be configured to screen target prompt information from the candidate prompt information according to the similarity and the traffic ratio corresponding to the candidate prompt information, and recommend the target prompt information to the user.

The input prompting device for the question and answer robot provided by the embodiment of the specification preprocesses the question input information received by the question and answer robot, selects and searches candidate prompting information in the knowledge point database based on the preprocessed information, and screens out target prompting information by taking text similarity and frequency information of the candidate prompting information, namely flow ratio, as a measurement index, so that the accuracy of screening the target prompting information is improved, and the target prompting information is closer to the requirements of users. And recommending the obtained target prompt information to the user for selection by the user, so that the workload of the user is reduced. Meanwhile, the recommended target prompt information generally conforms to the question format of the question-answering robot, so that the data processing amount of the question-answering robot is reduced, and the capability of the question-answering robot for accurately answering the user questions is improved.

On the basis of the foregoing embodiments, in some embodiments of the present specification, the candidate information obtaining module is further configured to:

In the embodiment of the specification, the candidate prompt information is searched from the knowledge point database, and the high-frequency question matched with the input question information can be searched from the high-frequency question database to serve as the candidate prompt information, so that the content of the prompt information is expanded, and the coverage rate and the accuracy of the recommended prompt information are improved.

On the basis of the foregoing embodiments, in some embodiments of the present specification, the preprocessing module is specifically configured to:

In the embodiment of the specification, the question information input by the user is synonymously expanded, so that the secondary high rate of the candidate prompt information is increased, and the accuracy of target prompt information recommendation is improved.

Fig. 4 is a schematic structural diagram of an input prompting device for a question-answering robot in another embodiment of the present specification, and as shown in fig. 4, on the basis of the above embodiment, in some embodiments of the present specification, the device further includes a knowledge point database construction module 41 configured to:

and configuring the knowledge point database for the question-answering robot in advance, wherein the knowledge point database comprises knowledge point titles and corresponding answers.

In the embodiment of the specification, a knowledge point database is configured for the question and answer robot, and when prompt information is recommended for a user by inputting question information, a matched knowledge point title is inquired based on the configured knowledge point database, so that prompt information is accurately recommended, and a function of quickly giving a corresponding answer is provided.

Fig. 5 is a schematic structural diagram of an input prompting device for a question-answering robot in another embodiment of the present specification, and as shown in fig. 5, on the basis of the above embodiment, in some embodiments of the present specification, the device further includes a high-frequency question bank constructing module 51, configured to:

On the basis of the foregoing embodiments, in some embodiments of the present specification, the prompt information filtering module is specifically configured to:

and sorting candidate prompt information with the score larger than a specified threshold or with the score sorted in a preset ranking from high to low as the target prompt information.

In the embodiment of the specification, candidate prompt information is converted into multi-dimensional feature vectors through barrel dividing operation, and then the candidate prompt information is graded and screened by utilizing a prompt information grading model to obtain target prompt information meeting conditions and recommended to a user, so that automatic prompt of input information of a question and answer robot is realized.

On the basis of the foregoing embodiments, in some embodiments of the present specification, the apparatus further includes a model optimization module, configured to:

and optimizing the prompt information scoring model according to the click data of the target prompt information.

In the embodiment of the description, the click data of the target prompt information is used as sample data, and the prompt information scoring model is optimally trained to improve the output accuracy of the prompt information scoring model and further improve the accuracy of the prompt information.

It should be noted that the above-described apparatus may also include other embodiments according to the description of the method embodiment. The specific implementation manner may refer to the description of the above corresponding method embodiment, and is not described in detail herein.

An embodiment of the present specification further provides an input prompt apparatus for a question and answer robot, including: at least one processor and a memory for storing processor-executable instructions, the processor implementing the input prompting method for the question-answering robot in the above embodiments when executing the instructions, such as:

determining the similarity between the candidate prompt information and the pre-processing question-raising information by using a text similarity algorithm;

acquiring the traffic ratio of the candidate prompt information in a specified time range;

An embodiment of the present specification further provides a robot for question answering, including: at least one processor and a memory for storing processor-executable instructions, the processor implementing the input prompting method for the question-answering robot in the above embodiments when executing the instructions, such as:

And the processor is also used for outputting a corresponding answer according to the clicking operation of the target prompt information by the user.

It should be noted that, the processing device and the question-answering robot described above may also include other embodiments according to the description of the method embodiment. The specific implementation manner may refer to the description of the above corresponding method embodiment, and is not described in detail herein.

The input prompting device or processing equipment for the question answering robot or the question answering robot provided by the specification can also be applied to various data analysis and processing systems. The system or apparatus or processing device may include any of the input prompting apparatuses for the question-answering robot in the above embodiments. The system or apparatus or processing device may be a single server, or may include a server cluster, a system (including a distributed system), software (applications), an actual operation device, a logic gate device, a quantum computer, etc. using one or more of the methods or one or more of the embodiments of the present disclosure, and a terminal device incorporating necessary hardware for implementation. The system for checking for discrepancies may comprise at least one processor and a memory storing computer-executable instructions that, when executed by the processor, implement the steps of the method of any one or more of the embodiments described above.

The method embodiments provided by the embodiments of the present specification can be executed in a mobile terminal, a computer terminal, a server or a similar computing device. Taking an example of the server running on the server, fig. 6 is a hardware structure block diagram of an input prompting server for the question-answering robot in an embodiment of the present specification, where the server may be the input prompting device for the question-answering robot, the input prompting apparatus for the question-answering robot, or the system in the above-mentioned embodiment. As shown in fig. 6, the server 10 may include one or more (only one shown) processors 100 (the processors 100 may include, but are not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 200 for storing data, and a transmission module 300 for communication functions. It will be understood by those skilled in the art that the structure shown in fig. 6 is merely illustrative and is not intended to limit the structure of the electronic device. For example, the server 10 may also include more or fewer components than shown in FIG. 6, and may also include other processing hardware, such as a database or multi-level cache, a GPU, or have a different configuration than shown in FIG. 6, for example.

The memory 200 may be used to store software programs and modules of application software, such as program instructions/modules corresponding to the input prompting method for the question-answering robot in the embodiment of the present specification, and the processor 100 executes various functional applications and resource data updates by executing the software programs and modules stored in the memory 200. Memory 200 may include high speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, memory 200 may further include memory located remotely from processor 100, which may be connected to a computer terminal through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The transmission module 300 is used for receiving or transmitting data via a network. Specific examples of the network described above may include a wireless network provided by a communication provider of the computer terminal. In one example, the transmission module 300 includes a Network adapter (NIC) that can be connected to other Network devices through a base station so as to communicate with the internet. In one example, the transmission module 300 may be a Radio Frequency (RF) module, which is used for communicating with the internet in a wireless manner.

The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.

The method or apparatus provided by the present specification and described in the foregoing embodiments may implement service logic through a computer program and record the service logic on a storage medium, where the storage medium may be read and executed by a computer, so as to implement the effect of the solution described in the embodiments of the present specification.

The storage medium may include a physical device for storing information, and typically, the information is digitized and then stored using an electrical, magnetic, or optical media. The storage medium may include: devices that store information using electrical energy, such as various types of memory, e.g., RAM, ROM, etc.; devices that store information using magnetic energy, such as hard disks, floppy disks, tapes, core memories, bubble memories, and usb disks; devices that store information optically, such as CDs or DVDs. Of course, there are other ways of storing media that can be read, such as quantum memory, graphene memory, and so forth.

The input prompting method or device for the question-answering robot provided in the embodiments of the present specification may be implemented in a computer by a processor executing corresponding program instructions, for example, implemented in a PC using a c + + language of a windows operating system, implemented in a linux system, or implemented in an intelligent terminal using android, iOS system programming languages, implemented in processing logic based on a quantum computer, and the like.

It should be noted that descriptions of the apparatus, the computer storage medium, and the system described above according to the related method embodiments may also include other embodiments, and specific implementations may refer to descriptions of corresponding method embodiments, which are not described in detail herein.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the hardware + program class embodiment, since it is substantially similar to the method embodiment, the description is simple, and the relevant points can be referred to only the partial description of the method embodiment.

The embodiments of the present description are not limited to what must be consistent with industry communications standards, standard computer resource data updating and data storage rules, or what is described in one or more embodiments of the present description. Certain industry standards, or implementations modified slightly from those described using custom modes or examples, may also achieve the same, equivalent, or similar, or other, contemplated implementations of the above-described examples. The embodiments using the modified or transformed data acquisition, storage, judgment, processing and the like can still fall within the scope of the alternative embodiments of the embodiments in this specification.

In the 90 s of the 20 th century, improvements in a technology could clearly distinguish between improvements in hardware (e.g., improvements in circuit structures such as diodes, transistors, switches, etc.) and improvements in software (improvements in process flow). However, as technology advances, many of today's process flow improvements have been seen as direct improvements in hardware circuit architecture. Designers almost always obtain the corresponding hardware circuit structure by programming an improved method flow into the hardware circuit. Thus, it cannot be said that an improvement in the process flow cannot be realized by hardware physical modules. For example, a Programmable Logic Device (PLD), such as a Field Programmable Gate Array (FPGA), is an integrated circuit whose Logic functions are determined by programming the Device by a user. A digital system is "integrated" on a PLD by the designer's own programming without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Furthermore, nowadays, instead of manually making an integrated Circuit chip, such Programming is often implemented by "logic compiler" software, which is similar to a software compiler used in program development and writing, but the original code before compiling is also written by a specific Programming Language, which is called Hardware Description Language (HDL), and HDL is not only one but many, such as abel (advanced Boolean Expression Language), ahdl (alternate Language Description Language), traffic, pl (core unified Programming Language), HDCal, JHDL (Java Hardware Description Language), langue, Lola, HDL, laspam, hardsradware (Hardware Description Language), vhjhd (Hardware Description Language), and vhigh-Language, which are currently used in most common. It will also be apparent to those skilled in the art that hardware circuitry that implements the logical method flows can be readily obtained by merely slightly programming the method flows into an integrated circuit using the hardware description languages described above.

The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, an Application Specific Integrated Circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic for the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller as pure computer readable program code, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may thus be considered a hardware component, and the means included therein for performing the various functions may also be considered as a structure within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.

The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. One typical implementation device is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a vehicle-mounted human-computer interaction device, a cellular telephone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

Although one or more embodiments of the present description provide method operational steps as described in the embodiments or flowcharts, more or fewer operational steps may be included based on conventional or non-inventive approaches. The order of steps recited in the embodiments is merely one manner of performing the steps in a multitude of orders and does not represent the only order of execution. When the device or the end product in practice executes, it can execute sequentially or in parallel according to the method shown in the embodiment or the figures (for example, in the environment of parallel processors or multi-thread processing, even in the environment of distributed resource data update). The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, the presence of additional identical or equivalent elements in a process, method, article, or apparatus that comprises the recited elements is not excluded. The terms first, second, etc. are used to denote names, but not any particular order.

For convenience of description, the above devices are described as being divided into various modules by functions, and are described separately. Of course, when implementing one or more of the present description, the functions of each module may be implemented in one or more software and/or hardware, or a module implementing the same function may be implemented by a combination of multiple sub-modules or sub-units, etc. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable resource data updating apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable resource data updating apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable resource data update apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable resource data update apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage, graphene storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

As will be appreciated by one skilled in the art, one or more embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, one or more embodiments of the present description may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, one or more embodiments of the present description may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

One or more embodiments of the present description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. One or more embodiments of the present specification can also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, and the relevant points can be referred to only part of the description of the method embodiments. In the description of the specification, reference to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the specification. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.

The above description is merely exemplary of one or more embodiments of the present disclosure and is not intended to limit the scope of one or more embodiments of the present disclosure. Various modifications and alterations to one or more embodiments described herein will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement or the like made within the spirit and principle of the present specification should be included in the scope of the claims.

Claims

1. An input prompting method for a question-answering robot comprises the following steps:

2. The method of claim 1, further comprising:

3. The method of claim 1, wherein the performing word segmentation and normalization on the received question input information to obtain pre-processed question information comprises:

4. The method of claim 1, further comprising:

5. The method of claim 2, further comprising:

6. The method according to claim 1, wherein the screening out target prompt information from the candidate prompt information according to the similarity and traffic ratio corresponding to the candidate prompt information comprises:

7. The method of claim 6, further comprising:

8. An input prompting device for a question-answering robot, comprising:

9. The apparatus of claim 8, the candidate information acquisition module further to:

10. The apparatus of claim 8, the pre-processing module to:

11. The apparatus of claim 8, the apparatus further comprising a knowledge point database building module to:

12. The apparatus of claim 9, further comprising a high frequency problem library building module to:

13. The apparatus of claim 8, wherein the hint information filter module is specifically configured to:

14. The apparatus of claim 13, further comprising a model optimization module to:

15. An input prompting device for a question-answering robot, comprising: at least one processor and a memory for storing processor-executable instructions, the processor implementing the method of any one of claims 1-7 when executing the instructions.

16. A question-answering robot comprising: at least one processor and a memory for storing processor-executable instructions, the processor implementing the method of any one of claims 1-7 when executing the instructions;