CN112905752A

CN112905752A - Intelligent interaction method, device, equipment and storage medium

Info

Publication number: CN112905752A
Application number: CN202110340049.1A
Authority: CN
Inventors: 王全礼; 李昱; 张晨; 蒲柯锐; 王斌; 邓尧文; 张圳
Original assignee: China Construction Bank Corp
Current assignee: China Construction Bank Corp
Priority date: 2021-03-30
Filing date: 2021-03-30
Publication date: 2021-06-04

Abstract

The application relates to the technical field of artificial intelligence and discloses an intelligent interaction method, device, equipment and storage medium. Acquiring at least two core words related to an instant message sent by a user, and determining the weight of the at least two core words and the incidence relation between the at least two core words; selecting a target question of the instant message from an existing question bank according to the at least two core words, the weights and the incidence relation of the at least two core words; and acquiring answer information associated with the target question as reply information of the instant message, and feeding back the reply information to the user. Through the technical scheme, the problem that the answer range of the traditional search engine is too large is solved, the matching accuracy of the answers is improved, the user experience is improved, and a new idea is provided for realizing intelligent interaction.

Description

Intelligent interaction method, device, equipment and storage medium

Technical Field

The embodiment of the application relates to the technical field of artificial intelligence, in particular to an intelligent interaction method, an intelligent interaction device, intelligent interaction equipment and a storage medium.

Background

With the rapid development of the internet and the artificial intelligence technology, the intelligent interaction plays an increasingly greater role in the aspects of user experience and information acquisition, the current intelligent question-answering and information retrieval methods are typical intelligent interaction modes, but the intelligent question-answering is strict in answer matching, the content given by the information retrieval is too wide, and a user needs a certain cost for acquiring required information from a retrieval result, so that a novel intelligent interaction method and a novel intelligent interaction flow need to be constructed, and the problem of accurately acquiring the information under different scenes of the user is solved.

At present, the intelligent question-answer data and the intelligent question-answer matching degree have high requirements, and no corresponding answer is obtained for the question of the user in many cases; secondly, in the prior art, the content of the result obtained by information retrieval is more and wider, and the user needs to spend time and cost for reading and filtering again, which affects the user experience.

Disclosure of Invention

The application provides an intelligent interaction method, an intelligent interaction device, intelligent interaction equipment and a storage medium, so that answer matching accuracy is improved, and user experience is improved.

In a first aspect, an embodiment of the present application provides an intelligent interaction method, including:

acquiring at least two core words related to an instant message sent by a user, and determining the weight of the at least two core words and the incidence relation between the at least two core words;

selecting a target question of the instant message from an existing question bank according to the at least two core words, the weights and the incidence relation of the at least two core words;

and acquiring answer information associated with the target question as reply information of the instant message, and feeding back the reply information to the user.

In a second aspect, an embodiment of the present application further provides an intelligent interaction apparatus, including:

the core information determining module is used for acquiring at least two core words related to the instant message sent by the user and determining the weight of the at least two core words and the incidence relation between the at least two core words;

the target question determining module is used for selecting a target question of the instant message from an existing question bank according to the at least two core words, the weight and the incidence relation of the at least two core words;

and the answer information acquisition module is used for acquiring answer information associated with the target question as answer information of the instant message and feeding back the answer information to the user.

In a third aspect, an embodiment of the present application further provides an electronic device, including:

one or more processors;

a memory for storing one or more programs;

when executed by the one or more processors, cause the one or more processors to implement an intelligent interaction method as provided by any embodiment of the application.

In a fourth aspect, this application further provides a computer-readable storage medium, on which a computer program is stored, where the computer program is executed by a processor to implement the intelligent interaction method as provided in any embodiment of this application.

According to the method and the device, at least two core words related to the instant message sent by the user are obtained, the weight of the at least two core words and the incidence relation between the at least two core words are determined, then the target question of the instant message is selected from an existing question bank according to the weight of the at least two core words and the incidence relation between the at least two core words, answer information related to the target question is obtained and serves as reply information of the instant message, and the reply information is fed back to the user. According to the technical scheme, the core words are obtained by fully analyzing the instant messages, the target problems related to the instant messages can be accurately positioned from the existing problem library by combining the weight, the incidence relation and the like of the core words, namely, the range of the reply information of the instant messages is defined.

Drawings

Fig. 1 is a flowchart of an intelligent interaction method provided in an embodiment of the present application;

fig. 2 is a flowchart of an intelligent interaction method provided in the second embodiment of the present application;

fig. 3 is a flowchart of an intelligent interaction method provided in the third embodiment of the present application;

fig. 4 is a flowchart of an intelligent interaction method provided in the fourth embodiment of the present application;

fig. 5 is a schematic structural diagram of an intelligent interaction device provided in the fifth embodiment of the present application;

fig. 6 is a schematic structural diagram of an electronic device according to a sixth embodiment of the present application.

Detailed Description

The present application will be described in further detail with reference to the following drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the application and are not limiting of the application. It should be further noted that, for the convenience of description, only some of the structures related to the present application are shown in the drawings, not all of the structures.

Example one

Fig. 1 is a flowchart of an intelligent interaction method provided in an embodiment of the present application; the method can be applied to the scenes of government affairs cloud, intelligent customer service, intelligent outbound, intelligent information and the like, can be executed by an intelligent interaction device, is realized by software/hardware, and can be integrated in electronic equipment bearing an intelligent interaction function, such as a server.

As shown in fig. 1, the method may specifically include:

s110, obtaining at least two core words related to the instant message sent by the user, and determining the weight of the at least two core words and the incidence relation between the at least two core words.

The instant message refers to a message input by a user on an interactive interface in the scenes of government affairs cloud, intelligent customer service, intelligent outbound, intelligent information and the like. The core word is a word for representing the core meaning in the instant message, such as product name, concept, service introduction, product operation, time, place, reference information, and the like, wherein the reference information refers to information obtained by combining historical conversation records. The weight of the core word is used for representing the importance degree of the core word in the instant message, and the larger the weight is, the more important the core word is in the instant message. The incidence relation between the core words is used for representing the degree of closeness between the core words, and can be a relation of or, a sequential relation, a dependency relation and the like.

In this embodiment, at least two core words may be extracted from the instant message. Specifically, the core words may be extracted from the instant message based on a word segmentation technique. Core words may also be extracted from instant messages based on entity extraction techniques. For example, the instant message sent by the user is "how do credit card is lost? ", the extracted core words are" credit card "," lost ".

Further, at least two core words can be extracted from the historical conversation record in the current scene associated with the instant message, so as to fully understand the purpose or intention of the user. Specifically, based on the dynamic content generator, at least two core words are extracted from the historical conversation record in the current scene associated with the instant message. For example, the instant message "how to complement the card" input by the user is associated with the historical conversation record "user: how is the credit card lost? The robot comprises: the credit card is reported lost. The user: how to complement the card, the core words extracted are credit card and complement card.

For example, as another optional way of the embodiment of the present application, in order to further improve the accuracy of obtaining the answer, the embodiment may preferably extract the core word from the instant message and the historical conversation record in the current scenario associated with the instant message at the same time.

Optionally, after determining at least two core words associated with the instant message, in order to facilitate better understanding of the instant message and to obtain reply information of the instant message more accurately, the present embodiment may determine a weight of each core word. For example, the weight of each core word may be determined in combination with the sentence structure (the principal predicate) of the instant message, the part of speech of the core word, and the like; and the necessity judgment can be carried out on the at least two core words based on the necessity model, so that the weights of the at least two core words are obtained. Specifically, the convolutional neural network model is trained by using an existing data set through inputting core words, parts of speech, word frequency, mutual information and the like in a sentence, and finally the weight of the core words in the sentence is output.

Optionally, the association relationship between at least two core words may also be determined according to the part of speech, mutual information, weight of the core words, and the like of the core words.

It can be understood that the core words associated with the instant messages and the association relation between the weight of the core words and the core words are obtained through the analysis of the instant messages, and a foundation is laid for accurately judging the target problems of the instant messages in the follow-up process.

S120, selecting a target question of the instant message from the existing question bank according to the at least two core words, the weight and the incidence relation of the at least two core words.

Wherein, the target question refers to the question which is matched with the instant message of the user in the existing questions.

In this embodiment, the target question of the instant message is selected from the existing question library based on a machine learning algorithm according to the at least two core words, the weight and the association relationship of the at least two core words.

Further, in order to accurately determine the target question associated with the instant message, the target question of the instant message may be selected from an existing question library according to at least two core words, the weight and the association relationship of the at least two core words, and the service scenario associated with the instant message. The service scene associated with the instant message refers to a scene of a service type consulted by the user, such as a credit card service, a deposit card service, and the like.

Specifically, the target question of the instant message can be selected from an existing question bank based on a machine learning algorithm according to at least two core words, the weight and the incidence relation of the at least two core words and the service scene associated with the instant message.

It can be understood that introducing the service scenario can more accurately understand the intention of the user, select the target problem of the user instant message from the existing problem library, and meet the user requirement.

It should be noted that, when a plurality of questions are matched from the existing question bank, the weights of the core words need to be updated, specifically, the necessity module may be retrained by adjusting the iteration number of the necessity model, and the like, and then the weights of at least two core words are obtained again according to the necessity module. Furthermore, the target problem can be determined again from the existing question bank according to the at least two newly acquired word core words, the weights and the incidence relations of the at least two core words.

S130, obtaining answer information related to the target question as reply information of the instant message, and feeding back the reply information to the user.

The answer information refers to answer information in an answer library which is established in advance and corresponds to the question, and the answer information may include at least one piece of information.

In this embodiment, answer information associated with the target question is obtained from the answer library, and specifically, the answer information corresponding to the target question may be found from the answer library by using the target question as an index. And then, the answer information is used as the reply information of the instant message of the user, and the reply information is fed back to the user. Furthermore, when the answer information is multiple pieces, the answer information can be assembled and sorted according to a certain mode, and the sorted answer information is used as the reply information of the instant message and is sequentially displayed to the user according to the sequence.

On the basis of the above technical solution, in order to more accurately understand the intention of the instant message of the user, as an optional way of this embodiment, after at least two core words associated with the instant message sent by the user are obtained, normalization processing is performed on the core words. Specifically, irrelevant words (such as passenger words, null words, nonsense words, and the like) are filtered out; if the instant message contains pinyin or the first letter translated into Chinese vocabulary, the instant message mainly refers to business words and common words, such as xinyongka (credit card), xyk (credit card); it is also possible to make alias or synonymous phrase supplements, such as debit cards (credit cards), what to call (name), etc.; wrongly written words can also be corrected, such as credit card, etc.

Example two

Fig. 2 is a flowchart of an intelligent interaction method provided in the second embodiment of the present application; on the basis of the above embodiment, an alternative scheme is provided for further optimizing the target problem of selecting the instant message from the existing problem library according to the weight and the incidence relation of at least two core words.

As shown in fig. 2, the method may specifically include:

s210, obtaining at least two core words related to the instant message sent by the user, and determining the weight of the at least two core words and the incidence relation between the at least two core words.

S220, selecting a target question of the instant message from the existing question bank according to the at least two core words, the weight and the incidence relation of the at least two core words and the service scene associated with the instant message.

In this embodiment, the target analysis information associated with the instant message may be determined according to the at least two core words, the weight and the association relationship of the at least two core words, and the service scenario associated with the instant message based on the set problem analysis template. The set problem analysis template is an analysis template of problem information associated with problems, each problem corresponds to one set problem analysis template, the set problem analysis template comprises analysis information, the analysis information comprises at least one of basic attributes, associated attributes and extended attributes, and the basic attributes comprise core words, weight of the core words, service scenes and the like; the association attribute comprises association relation between core words, such as A, B two core words, A | B represents that A and B have no sequence relation, A- > B represents that A and B have sequence relation, and AB represents that A and B have dependency relation; the extended attribute is the extension of the core word, such as semantic related information, synonymous information, etc., i.e. the core word extracted from the historical conversation record in the current scene of the instant message.

Specifically, at least two core words obtained from the instant message, the weight and the association relationship of the at least two core words, and the service scene associated with the instant message may be respectively filled in a set problem analysis template, so as to obtain target analysis information associated with the instant message.

And further, after the target analysis information related to the instant message is determined, selecting the target question of the instant message from the existing question bank according to the matching result between the target analysis information and the analysis information related to the existing question in the existing question bank.

Specifically, the target analysis information may be used as an index, the existing problem corresponding to the analysis information associated with the existing problem in the existing problem library matched with the target analysis information may be searched in the existing problem library, and the existing problem may be used as the target problem of the instant message.

Exemplarily, the similarity between the target analysis information and analysis information associated with existing problems in an existing problem library can be determined; and selecting the target question of the instant message from the existing question bank according to the similarity. Specifically, the similarity between the target analysis information and the analysis information associated with the existing problems in the existing problem library is calculated, and the existing problem associated with the analysis information corresponding to the maximum similarity is selected from the similarities larger than the set threshold value and is used as the target problem of the instant message. If the similarity is smaller than the set threshold, the target analysis information related to the instant message is not accurate enough, and the target analysis information related to the instant message is determined again. Wherein, the setting of the threshold is set by the person skilled in the art according to the actual situation.

It can be understood that the speed of determining the target question of the instant message can be improved by introducing the similarity and determining the associated question of the instant message from the existing questions in the existing question bank.

And S230, acquiring answer information associated with the target question as reply information of the instant message, and feeding back the reply information to the user.

According to the technical scheme of the embodiment, target analysis information associated with the instant message is determined according to at least two core words, the weight and the association relation of the at least two core words and a service scene associated with the instant message based on a set problem analysis template, and then the target problem of the instant message is selected from an existing problem library according to a matching result between the target analysis information and the analysis information associated with the existing problems in the existing problem library. A set problem analysis template is introduced to assist in determining target analysis information associated with the instant message, so that the target problem of the instant message is determined, the accuracy of determining the target problem is improved, and further guarantee is provided for obtaining answers to subsequent target problems.

EXAMPLE III

Fig. 3 is a flowchart of an intelligent interaction method provided in the third embodiment of the present application; on the basis of the above embodiment, further optimization is performed on the 'obtaining answer information associated with the target problem', and an optional implementation scheme is provided.

As shown in fig. 3, the method may specifically include:

s310, obtaining at least two core words related to the instant message sent by the user, and determining the weight of the at least two core words and the association relation between the at least two core words.

S320, selecting the target question of the instant message from the existing question bank according to the at least two core words, the weight and the incidence relation of the at least two core words.

S330, determining a target calculation engine from the optional calculation engines according to the target problem of the instant message.

The computing engine is a tool for acquiring answers to Questions, and the commonly used computing engines include professional computing engines, which mainly include four categories of TableQA (natural language query table data tool), knowledge base Question Answering (KB-QA), FAQ retrieval systems (FAQ), and Reading Comprehension-based Question Answering (MRCQA). Wherein, TableQA refers to automatically generating SQL statements, that is, NLP2SQL, according to the intention understanding template, and querying the database for answers, for example, how many users there are credit cards, which are solved by TableQA; KB-QA refers to knowledge graph-based QA, which obtains answers through graph computation and reasoning capabilities, such as what types of credit cards are; the FAQ has a question-answer function, such as a credit card transaction function and other navigation functions; MRCQA refers to QA based on reading understanding, such as common sense of credit cards, by which common sense key information related to credit cards can be found from documents related to credit cards. The computational engines for answer acquisition for different questions are different. The optional calculation engine is a calculation engine which is selected currently; the target calculation engine refers to a tool required for obtaining an answer corresponding to the target question, i.e., one of the calculation engines can be selected.

In this embodiment, the target problem of the instant message is used as an index, and the target problem is sequentially matched with the problem associated with each computing engine in the data source, and if the matching is successful, the computing engine associated with the problem is used as the target computing engine. The data source includes information of existing questions, business functions, optional calculation engines, etc., and is determined by those skilled in the art according to a large number of questions and calculation engines for obtaining answers to the questions.

Optionally, as an optional manner of the embodiment of the application, a candidate calculation engine may be determined from the optional calculation engines according to a target problem of the instant message, and then a target calculation engine is determined according to the candidate calculation engine.

Specifically, candidate calculation engines can be determined from the optional calculation engines according to the target problem of the instant message based on the deep learning technology, and then the target calculation engine can be determined according to the number of various calculation engines in the candidate calculation engines. For example, the number of tableqas in the candidate computing engine is 2; the number of KB-QA is 1; the number of FAQs is 3; and if the number of the MRCQAs is 6, taking the MRCQAs as a target computing engine.

S340, obtaining answer information related to the target question through the target calculation engine.

In this embodiment, answer information associated with the target question is obtained by the target calculation engine.

And S350, taking the answer information as the reply information of the instant message, and feeding back the reply information to the user.

According to the technical scheme, the core words are obtained by fully analyzing the instant messages, the target problems related to the instant messages can be accurately positioned from the existing problem library by combining the weight and the incidence relation of the core words, namely, the range of the reply information of the instant messages is defined.

Example four

Fig. 4 is a flowchart of an intelligent interaction method provided in the fourth embodiment of the present application; on the basis of the above embodiment, an alternative implementation scheme is provided for optimizing "determining a target computing engine from alternative computing engines according to a target problem of instant messaging".

As shown in fig. 4, the method may specifically include:

s410, obtaining at least two core words related to the instant message sent by the user, and determining the weight of the at least two core words and the incidence relation between the at least two core words.

S420, selecting a target question of the instant message from the existing question bank according to the at least two core words, the weight and the incidence relation of the at least two core words.

S430, according to the target problem of the instant message, candidate calculation engines are determined from the optional calculation engines.

In this embodiment, according to the target problem of the instant message, the candidate calculation engines may be determined from the optional calculation engines by extracting keywords from the target problem of the instant message, using the keywords as an index, and determining the candidate calculation engines from the optional calculation engines based on the constructed keyword inverted index.

The keywords are words used for representing core meanings in the target problem, and can also be core words associated with instant messages. The keyword inverted index is constructed in advance by the technicians in the field according to the existing question bank, question answers and the like. Illustratively, the keyword inverted index may be constructed by: extracting keywords from problem information associated with optional computing engines of the data source based on a word segmentation technology; and constructing the keyword inverted index according to the incidence relation between the keywords and the optional calculation engine. The data source includes question information, optional calculation engine, service function, etc. The keyword inverted index comprises a keyword retrieval domain and an engine number, and the corresponding engine number can be quickly found by retrieving the keyword in the keyword retrieval domain, so that the corresponding calculation engine is determined, wherein the engine number is associated with the calculation engine.

Specifically, keywords can be extracted from the target question of the instant message based on the word segmentation technology. Further, the keywords are used as indexes, and the constructed keyword reverse indexes are input to obtain at least one computing engine; and searching whether the obtained at least one computing engine exists from the optional computing engines, and if so, taking the obtained at least one computing engine as a candidate computing engine.

Further, the present embodiment may also construct a keyword inverted index according to the data source and the historical question and answer data. Specifically, a first keyword retrieval domain of the keyword inverted index is constructed according to the data source, and then a second keyword retrieval domain of the keyword inverted index is constructed according to the historical question-answer data. Extracting keywords from each piece of data in the historical question-answer data, and constructing a second keyword retrieval domain of the keyword inverted index according to the incidence relation between the keywords and the answer engine.

Then, the keywords are respectively used as indexes, retrieval is carried out in a first keyword retrieval domain in the constructed keyword inverted index to obtain at least one computing engine, whether the obtained at least one computing engine exists or not is searched from the selectable computing engines, and if the obtained at least one computing engine exists, the obtained at least one computing engine is used as a first candidate computing engine; and searching a second keyword search domain in the constructed keyword inverted index to obtain at least one computing engine, searching whether the obtained at least one computing engine exists in the optional computing engines, and if so, taking the obtained at least one computing engine as a second candidate computing engine. And taking the intersection of the first candidate calculation engine and the second candidate calculation engine as a candidate calculation engine. Optionally, a union of the first candidate compute engine and the second candidate compute engine may be used as the candidate compute engine.

And S440, determining a target calculation engine according to the candidate calculation engines.

In this embodiment, the target computing engine may be determined according to the priority of the candidate computing engine, the similarity between the target problem and the problem information associated with the candidate computing engine, and the number of various computing engines in the candidate computing engine. Wherein the greater the priority of the candidate compute engine, the greater the score of the corresponding compute engine.

Specifically, the target calculation engine is determined according to the priority of the candidate calculation engine, the similarity between the target problem and the problem information associated with the candidate calculation engine, the number of various calculation engines in the candidate calculation engine, and the set weight. For example, the number of each class of compute engines in the candidate compute engines, the similarity between the target problem and the problem information associated with that class of compute engines, the priority scores for that class of compute engines may be determined; and multiplying the number, the similarity and the priority score of the calculation engines with a set threshold value to obtain an engine score, and taking the calculation engine with the highest engine score of various calculation engines in the candidate calculation engines as a target calculation engine.

S450, obtaining answer information related to the target question through the target calculation engine.

And S460, using the answer information as the reply information of the instant message, and feeding back the reply information to the user.

According to the technical scheme, the candidate calculation engines are determined by introducing the inverted index, the target calculation engine is further determined, the determination speed of the calculation engines is increased, the determination speed of answers is further increased, and the user experience is improved.

EXAMPLE five

Fig. 5 is a schematic structural diagram of an intelligent interaction device provided in the fifth embodiment of the present application; the embodiment can be applied to the scenes of government affairs cloud, intelligent customer service, intelligent outbound, intelligent information and the like, and the device is realized by software/hardware and can be integrated in electronic equipment bearing an intelligent interaction function, such as a server.

As shown in fig. 5, the apparatus includes a core information determining module 510, a target question determining module 520, and an answer information acquiring module 530, wherein,

a core information determining module 510, configured to obtain at least two core words associated with an instant message sent by a user, and determine a weight of the at least two core words and an association relationship between the at least two core words;

a target question determining module 520, configured to select a target question of an instant message from an existing question bank according to at least two core words, and the weights and the association relationships of the at least two core words;

the answer information obtaining module 530 is configured to obtain answer information associated with the target question as answer information of the instant message, and feed back the answer information to the user.

Further, the core information determining module 510 includes a core word determining unit, which is specifically configured to:

extracting at least two core words from the instant message; and/or the presence of a gas in the gas,

at least two core words are extracted from a historical conversation record in a current scene associated with the instant message.

Further, the target problem determination module 520 is specifically configured to:

and selecting a target question of the instant message from the existing question library according to the at least two core words, the weight and the incidence relation of the at least two core words and the service scene associated with the instant message.

Further, the target problem determination module 520 includes a target resolution information determination unit and a target problem determination unit, wherein,

the target analysis information determining unit is used for determining target analysis information associated with the instant message according to the at least two core words, the weight and the association relation of the at least two core words and the service scene associated with the instant message based on the set problem analysis template;

and the target problem determining unit is used for selecting the target problem of the instant message from the existing problem library according to the matching result between the target analysis information and the analysis information related to the existing problems in the existing problem library.

Further, the target problem determination unit includes a similarity determination subunit and a target problem determination subunit, wherein,

the similarity determining subunit is used for determining the similarity between the target analysis information and analysis information associated with the existing problems in the existing problem library;

and the target question determining subunit is used for selecting the target question of the instant message from the existing question library according to the similarity.

Further, the answer information acquisition module 530 includes a target calculation engine determination sub-module and an answer information acquisition sub-module, wherein,

the target calculation engine determining submodule is used for determining a target calculation engine from the optional calculation engines according to the target problem of the instant message;

and the answer information acquisition submodule is used for acquiring the answer information related to the target question through the target calculation engine.

Further, the target calculation engine determination sub-module includes a candidate calculation engine determination unit and a target calculation engine determination unit, wherein,

the candidate calculation engine determining unit is used for determining candidate calculation engines from the optional calculation engines according to the target problem of the instant message;

and the target calculation engine determining unit is used for determining a target calculation engine according to the candidate calculation engines.

Further, the candidate calculation engine determination unit includes a keyword extraction sub-unit and a candidate calculation engine determination sub-unit, wherein,

a keyword extraction subunit, configured to extract keywords from a target question of an instant message;

and the candidate calculation engine determining subunit is used for determining candidate calculation engines from the optional calculation engines based on the constructed keyword inverted index by taking the keyword as the index.

Further, the candidate calculation engine determining unit further includes an inverted index determining subunit, where the inverted index determining subunit is specifically configured to:

extracting keywords from question information associated with optional computing engines of the data source;

and constructing the keyword inverted index according to the incidence relation between the keywords and the optional calculation engine.

Further, the inverted index determination subunit is further configured to:

and constructing a keyword inverted index according to the data source and the historical question and answer data.

Further, the target calculation engine determining unit is specifically configured to:

and determining the target computing engine according to the priority of the candidate computing engine, the similarity between the target problem and the problem information associated with the candidate computing engine and the number of various computing engines in the candidate computing engine.

Further, the core information determining module 510 further includes a weight determining unit, which is specifically configured to:

and based on the necessity model, performing necessity judgment on the at least two core words to obtain the weights of the at least two core words.

The intelligent interaction device can execute the intelligent interaction method provided by any embodiment of the application, and has corresponding functional modules and beneficial effects of the execution method.

EXAMPLE six

Fig. 6 is a schematic structural diagram of an electronic device according to a sixth embodiment of the present application, and fig. 6 shows a block diagram of an exemplary device suitable for implementing an embodiment of the present application. The device shown in fig. 6 is only an example, and should not bring any limitation to the function and the scope of use of the embodiments of the present application.

As shown in FIG. 6, electronic device 12 is embodied in the form of a general purpose computing device. The components of electronic device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 that couples various system components including the system memory 28 and the processing unit 16.

Bus 18 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a processor or local bus using any of a variety of bus architectures. By way of example, such architectures include, but are not limited to, Industry Standard Architecture (ISA) bus, micro-channel architecture (MAC) bus, enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

Electronic device 12 typically includes a variety of computer system readable media. Such media may be any available media that is accessible by electronic device 12 and includes both volatile and nonvolatile media, removable and non-removable media.

The system memory 28 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM)30 and/or cache memory 32. The electronic device 12 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, storage system 34 may be used to read from and write to non-removable, nonvolatile magnetic media (not shown in FIG. 6, and commonly referred to as a "hard drive"). Although not shown in FIG. 6, a magnetic disk drive for reading from and writing to a removable, nonvolatile magnetic disk (e.g., a "floppy disk") and an optical disk drive for reading from or writing to a removable, nonvolatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 by one or more data media interfaces. System memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the application.

A program/utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which examples or some combination thereof may comprise an implementation of a network environment. Program modules 42 generally perform the functions and/or methodologies of the embodiments described herein.

Electronic device 12 may also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), with one or more devices that enable a user to interact with electronic device 12, and/or with any devices (e.g., network card, modem, etc.) that enable electronic device 12 to communicate with one or more other computing devices. Such communication may be through an input/output (I/O) interface 22. Also, the electronic device 12 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the Internet) via the network adapter 20. As shown, the network adapter 20 communicates with other modules of the electronic device 12 via the bus 18. It should be understood that although not shown in the figures, other hardware and/or software modules may be used in conjunction with electronic device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.

The processing unit 16 executes various functional applications and data processing by executing programs stored in the system memory 28, for example, to implement the intelligent interaction method provided by the embodiment of the present application.

EXAMPLE seven

The seventh embodiment of the present application further provides a computer-readable storage medium, on which a computer program (or referred to as computer-executable instructions) is stored, where the computer program is used for executing the intelligent interaction method provided by the embodiment of the present application when the computer program is executed by a processor.

The computer storage media of the embodiments of the present application may take any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

Computer program code for carrying out operations for embodiments of the present application may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + +, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).

It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present application and the technical principles employed. It will be understood by those skilled in the art that the present application is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the application. Therefore, although the embodiments of the present application have been described in more detail through the above embodiments, the embodiments of the present application are not limited to the above embodiments, and many other equivalent embodiments may be included without departing from the spirit of the present application, and the scope of the present application is determined by the scope of the appended claims.

Claims

1. An intelligent interaction method, comprising:

2. The method of claim 1, wherein obtaining at least two core words associated with the instant message sent by the user comprises:

3. The method of claim 1, wherein selecting the target question of the instant message from an existing question bank according to the at least two core words, the weights and the association relationship of the at least two core words comprises:

and selecting the target question of the instant message from an existing question bank according to the at least two core words, the weight and the incidence relation of the at least two core words and the service scene associated with the instant message.

4. The method of claim 3, wherein selecting the target question of the instant message from an existing question bank according to the at least two core words, the weight and the association relationship of the at least two core words, and the service scenario associated with the instant message comprises:

based on a set problem analysis template, determining target analysis information associated with the instant message according to the at least two core words, the weight and the association relation of the at least two core words and the service scene associated with the instant message;

and selecting the target question of the instant message from the existing question bank according to a matching result between the target analysis information and analysis information associated with the existing question in the existing question bank.

5. The method of claim 4, wherein selecting the target question of the instant message from an existing question bank according to a matching result between the target parsing information and parsing information associated with existing questions in the existing question bank comprises:

determining the similarity between the target analysis information and analysis information associated with existing problems in an existing problem library;

and selecting the target question of the instant message from the existing question bank according to the similarity.

6. The method of claim 1, wherein obtaining answer information associated with the target question comprises:

determining a target computing engine from optional computing engines according to the target problem of the instant message;

and acquiring answer information associated with the target question through the target calculation engine.

7. The method of claim 6, wherein determining a target compute engine from among alternative compute engines based on the target question of the instant message comprises:

determining candidate computing engines from the optional computing engines according to the target problem of the instant message;

and determining a target calculation engine according to the candidate calculation engines.

8. The method of claim 7, wherein determining candidate compute engines from the alternative compute engines based on the target question of the instant message comprises:

extracting keywords from the target question of the instant message;

and determining candidate calculation engines from the optional calculation engines based on the constructed keyword reverse index by taking the keywords as indexes.

9. The method of claim 8, further comprising:

and constructing a keyword reverse index according to the incidence relation between the keywords and the optional calculation engine.

10. The method of claim 8, further comprising:

11. The method of claim 7, wherein determining a target compute engine from the candidate compute engines comprises:

12. The method of claim 1, wherein determining the weight of the at least two core words comprises:

13. An intelligent interaction device, comprising:

14. An electronic device, comprising:

one or more processors;

a memory for storing one or more programs;

the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the intelligent interaction method of any of claims 1-12.

15. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the intelligent interaction method according to any one of claims 1 to 12.