CN112148848A - A question and answer processing method and device - Google Patents

Abstract

The invention discloses a question and answer processing method and device, and relates to the technical field of artificial intelligence. One embodiment of the method comprises: acquiring a voice request; processing the voice request to obtain an intention classification corresponding to the voice request; and executing answer searching operation corresponding to the intention classification on the voice request, and feeding back a searching answer. Therefore, the voice requests with multiple intentions and incomplete intentions can be recognized, the active inquiry function is provided, and the intellectualization of the dialogue system is improved.

Description

A question and answer processing method and device

technical field

The present invention relates to the technical field of artificial intelligence, and in particular, to a question and answer processing method and device.

Background technique

Existing dialogue systems include a speech recognition module, a natural language understanding module, a dialogue management module, a language generation module and a speech generation module. Existing language recognition modules can only recognize single-intent voice requests, but cannot recognize multi-intent or incomplete-intent voice requests. For example, the user's voice request is "I want to listen", and the answer of the existing dialogue system is "preparing to play "Friends", "playing music for you" or "can't find this song", etc. For multi-intent voice requests, for example, the user's voice request is "play Harry Potter", and the existing dialogue system's answer is "ready to play a Harry Potter song", "play music for you" or "can't find this song" "Wait.

For voice requests with multiple intents or incomplete intents, the processing method of the existing dialogue system is relatively simple and rude. It directly divides multiple intents or incomplete intents into a single intent with a higher probability, and then returns the corresponding voice request according to the single intent. Answer. Although this process completes a round of dialogue, it does not fully reflect the initiative of the dialogue system, thus reducing the accuracy of the dialogue system's response.

SUMMARY OF THE INVENTION

In view of this, the embodiments of the present invention provide a question and answer processing method and device, which can identify voice requests with multiple intentions and incomplete intentions, and provide the function of active inquiry, which improves the intelligence of the dialogue system.

In order to achieve the above object, according to the first aspect of the embodiments of the present invention, there is provided a question and answer processing method, the method includes: obtaining a voice request; processing the voice request to obtain an intent classification corresponding to the voice request; The voice request performs an answer search operation corresponding to the intent classification, and feeds back the search answer.

Optionally, the processing of the voice request to obtain an intent classification corresponding to the voice request includes: performing semantic analysis on the voice request; using a model to perform intent recognition on the voice request; The voice request whose result is no scene and the intent recognition result is ambiguous intent is determined as incomplete intent; and the voice request whose semantic parsing result is scene and the intent recognition result is clear intent is determined as complete intent.

Optionally, performing an answer search operation corresponding to the intent classification on the voice request includes: if the intent classification is a complete intent, searching for the voice request in a scenario indicated by the semantic analysis result. If the intent classification is an incomplete intent, predict multiple preselected scenarios corresponding to the voice request, and select a preselected scenario that satisfies the first preset condition from the multiple preselected scenarios as candidate scenarios; send For the query request corresponding to the candidate scene, the answer to the voice request is searched in the candidate scene indicated by the query request result.

Optionally, the searching for the answer to the voice request in the scenario indicated by the semantic parsing result includes: if the semantic parsing result has an entity value, searching for the answer to the voice request in the scenario indicated by the semantic parsing result. and select the answer of the voice request from the search answer; if there is no entity value in the semantic parsing result, search for the answer corresponding to the scene indicated by the semantic parsing result, and select the answer from the search answer to select the answer to the voice request.

Optionally, if the semantic parsing result has an entity value, searching for an answer corresponding to the entity value in the scenario indicated by the semantic parsing result, including: if there is an entity value, then in the semantic Search for the answer corresponding to the entity value in the scenario indicated by the parsing result; if there are multiple entity values, determine whether the multiple entity values satisfy the same semantic slot at the same time; if so, in the scenario indicated by the semantic parsing result Next, the answer is searched for each of the entity values, and any answer is selected from the answers corresponding to each of the entity values; for a plurality of the entity values, the selected answers are spliced to obtain the The search answer corresponding to the entity value; if not, search for an answer that simultaneously satisfies a plurality of the entity values in the scenario indicated by the semantic parsing result.

Optionally, the first preset condition refers to a preselected scene with the top two confidence levels and a search answer.

In order to achieve the above object, according to the second aspect of the embodiments of the present invention, there is also provided a question and answer processing device, the processing device includes: an acquisition module for acquiring a voice request; a processing module for processing the voice request to obtain Intention classification corresponding to the voice request; a search module for performing an answer search operation corresponding to the intention classification to the voice request, and feeding back the search answer.

Optionally, the processing module includes: a semantic parsing unit for performing semantic parsing on the voice request; an intent recognition unit for performing intent recognition on the voice request by using a model; a determining unit for parsing the semantics The voice request whose result is no scene and the intent recognition result is ambiguous intent is determined as incomplete intent; and the voice request whose semantic parsing result is scene and the intent recognition result is clear intent is determined as complete intent.

Optionally, the search module includes: a first search unit, configured to search for an answer to the voice request in the scenario indicated by the semantic analysis result if the intent classification is a complete intent; a second search unit, If the intent classification is an incomplete intent, predict multiple pre-selected scenarios corresponding to the voice request, select a pre-selected scenario that satisfies the first preset condition from the multiple pre-selected scenarios as candidate scenarios, and send the For the query request corresponding to the candidate scene, the answer to the voice request is searched in the candidate scene indicated by the query request result.

Optionally, the first search unit includes: an entity value unit, configured to search for an answer corresponding to the entity value in the scenario indicated by the semantic analysis result if there is an entity value in the semantic analysis result, and Select the answer of the voice request from the search answer; the scene unit is used to search for the answer corresponding to the scene indicated by the semantic parsing result if the semantic parsing result does not have an entity value, and select the answer from the search answer answer the voice request.

To achieve the above object, according to a third aspect of the embodiments of the present invention, a computer-readable medium is further provided, on which a computer program is stored, and when the program is executed by a processor, the question and answer processing method described in the first aspect is implemented.

In the embodiment of the present invention, an intent classification corresponding to the voice request is obtained by processing the acquired voice request, and then an answer search operation corresponding to the intent classification is performed on the voice request, and the search answer is fed back. As a result, voice requests with multiple intentions and incomplete intentions can be recognized, and the function of active inquiry can be provided, which improves the intelligence of the dialogue system.

Further effects of the above non-conventional alternatives will be described below in conjunction with specific embodiments.

Description of drawings

The accompanying drawings are used for better understanding of the present invention and do not constitute an improper limitation of the present invention. In the drawings, the same or corresponding reference numerals denote the same or corresponding parts.

1 is a flowchart of a question and answer processing method according to an embodiment of the present invention;

2 is a flowchart of a question and answer processing method for an incomplete intent voice request according to an embodiment of the present invention;

3 is a flowchart of a question and answer processing method for a complete intent voice request according to an embodiment of the present invention;

4 is a schematic diagram of a question and answer processing apparatus according to an embodiment of the present invention;

FIG. 5 is an exemplary system architecture diagram to which an embodiment of the present invention may be applied;

FIG. 6 is a schematic structural diagram of a computer system suitable for implementing a terminal device or a server according to an embodiment of the present invention.

Detailed ways

Exemplary embodiments of the present invention are described below with reference to the accompanying drawings, which include various details of the embodiments of the present invention to facilitate understanding and should be considered as exemplary only. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the invention. Also, descriptions of well-known functions and constructions are omitted from the following description for clarity and conciseness.

As shown in FIG. 1, it is a flowchart of a question and answer processing method according to an embodiment of the present invention, and the method at least includes the following operation flow:

S101, acquiring a voice request.

Exemplarily, the voice request is obtained by means of text input or voice input.

S102: Process the voice request to obtain an intent classification corresponding to the voice request.

Exemplarily, perform semantic parsing on the voice request; use the model to perform intent recognition on the voice request; determine the voice request whose semantic parsing result is no scene and the intent recognition result is an ambiguous intent as an incomplete intent; and the semantic parsing result is A voice request that has a scene and the intent recognition result is a clear intent is determined to be a complete intent. When the following two situations occur in the processing result of the voice request, both belong to abnormal situations. When encountering such abnormal situations, it is usually necessary to optimize the semantic parsing module or the intent recognition module. The two cases are that the semantic parsing result is no scene and the intent recognition result is a clear intent, and the semantic parsing result is a scene and the intent recognition result is an ambiguous intent.

S103, perform an answer search operation corresponding to the intent classification on the voice request, and feed back the search answer.

Exemplarily, intent categories include complete intents and incomplete intents. If the intent classification of the voice request is an incomplete intent, multiple candidate answers corresponding to the voice request are searched, and an inquiry request is sent for the candidate answers to determine an answer to the voice request.

If the intent classification of the voice request is a complete intent, search for the answer to the voice request in the scenario indicated by the semantic parsing result; further, if there is an entity value in the semantic parsing result, search for the corresponding entity value in the scenario indicated by the semantic parsing result answer, and select the answer of the voice request from the search answer; if there is no entity value in the semantic parsing result, search for the answer corresponding to the scene indicated by the semantic parsing result, and select the answer of the voice request from the search answer.

Therefore, the embodiment of the present invention determines the intent classification of the voice request through the semantic analysis result and the intent recognition result, so as to perform an answer search operation corresponding to the intent classification on the voice request, thereby improving the intelligence of the dialogue system.

As shown in FIG. 2, it is a flowchart of a question and answer processing method for an incomplete intent voice request according to an embodiment of the present invention; the method at least includes the following operation flow:

S201, acquiring a voice request.

S202: Perform semantic analysis and intent recognition on the voice request, respectively, to obtain an intent corresponding to the voice request and classify it as an incomplete intent.

Exemplarily, the semantic analysis result of the voice request is no scene, and the intent recognition result of the voice request by the intent recognition model is an unclear intent; thus, it is determined that the intent corresponding to the voice request is classified as an incomplete intent.

S203: Predict multiple preselected scenarios corresponding to the voice request.

S204: Select a preselected scene that satisfies the first preset condition from the plurality of preselected scenes as a candidate scene.

Exemplarily, the first preset condition refers to a preselected scene with the top two confidence levels and a search answer. When the semantic parsing module predicts multiple pre-selected scenarios corresponding to the voice request, it outputs the confidence level of the pre-selected scenarios.

Here, the first preset condition is artificially set according to the actual application scenario.

S205, sending a query request corresponding to the candidate scene.

S206: Search for an answer to the voice request in the candidate scene indicated by the query request result.

Specifically, the voice request is "I want to listen", semantic analysis is performed on the voice request, and the semantic analysis result indicates that the voice request has no scene; the intent recognition model is used to identify the voice request, and the intent recognition result indicates that the voice request is The intent is ambiguous, whereby the voice request is determined to be an incomplete intent. Predict multiple pre-selected scenarios corresponding to the voice request through the semantic parsing module, output the confidence level of each pre-selected scenario, and select the pre-selected scenarios with the top two confidence scores and search answers from the multiple pre-selected scenarios, such as multiple pre-selected scenarios. The pre-selected scenes are sorted in descending order of the confidence of the pre-selected scenes, followed by music scene, poetry scene, audiobook scene, cross talk scene and opera scene, and the first two pre-selected scenes are music scene and poetry scene respectively; It is then determined whether the voice request has an answer in the context of music or poetry. If there is a search answer only in the music scene, for example, the search answers are "friend", "same song" and "lover", the music scene is used as the candidate scene for the voice request. If there is a search answer in both the music scene and the poetry scene, for example, the search answer in the poetry scene is "quatrain". Send a query request, the query request is "do you want to listen to music or poetry", if the user feedback content is "want to listen to music", select one of the corresponding search answers in the music scene as the answer to the voice request, such as from " "friend", "same song" and "lover", select "friend" as the answer of "I want to hear".

Here, the first search answer may be selected as the answer to the voice request based on the popularity of the search answer, or a search answer may be randomly selected as the answer to the voice request.

The embodiment of the present invention can identify the incomplete intention and provide the function of active inquiry for the incomplete intention, which realizes the initiative and intelligence of the dialogue system.

As shown in FIG. 3, it is a flowchart of a question and answer processing method for a complete intent voice request according to an embodiment of the present invention; the method includes at least the following operation process:

S301, acquiring a voice request.

S302: Perform semantic analysis and intent recognition on the voice request respectively, and obtain the intent corresponding to the voice request and classify it as a complete intent; if there is one entity value in the semantic analysis result, perform the operation of S303; if the semantic analysis result has multiple entity values, then Execute operation S304; if there is no entity value in the semantic analysis result, execute operation S307.

S303, search for an answer corresponding to the entity value in the scenario indicated by the semantic analysis result; then perform the operation of S308.

Exemplarily, there may be one or multiple search answers.

S304, determine whether the multiple entity values satisfy the same semantic slot at the same time; if so, execute the operation S305, and if not, execute the operation S306.

S305, in the scenario indicated by the semantic analysis result, search for answers for each entity value, and select any answer from the answers corresponding to each entity value; for multiple entity values, splicing the selected answers to obtain The search answer corresponding to the entity value; then perform the S308 operation.

S306, search for an answer that satisfies multiple entity values at the same time in the scenario indicated by the semantic analysis result; then perform the operation of S308.

S307, search for an answer corresponding to the scene indicated by the semantic analysis result; then perform the operation of S308.

S308, select the answer of the voice request from the search answers.

Exemplarily, if there is a search answer, the search answer is used as the answer to the voice request. If there are multiple search answers, a search answer that satisfies the preset condition is selected from the multiple search answers as the answer to the voice request.

Specifically, the voice request is "I want to listen to my friend's song", the semantic analysis result indicates that the scene of the voice request is "song" and the entity value is "friend", and the intent recognition result is a clear intent, so it is determined that the voice request is Intents are classified as full intents. Since the voice request has an entity value, search for the answer corresponding to "friend" in the "song" scene, and get three search answers, such as "friend" sung by Jay Chou, "friend" sung by Na Ying, and "friend" sung by Gao Gao "friend". According to the popularity of "friends" sung by different singers, the first one is selected as the answer to the voice request, or a search answer is randomly selected as the answer to the voice request.

The voice request is "I want to listen to music", the semantic parsing result indicates that the scene of the voice request is "music" and there is no entity value, and the intent recognition result is a clear intent, so it is determined that the intent of the voice request is classified as a complete intent. Since the voice request does not have an entity value, the answer is searched in the "music" scene, and three search answers are obtained, namely "friend", "same song" and "lover". Select the most popular search answer from the three search answers as the answer to the voice request, for example, select "friend" as the answer to "I want to listen to music".

The voice request is "check the weather in Beijing and Shanghai for me", the semantic parsing result indicates that the scene of the voice request is "weather", and the entity values are "Beijing" and "Shanghai" respectively, and the intent recognition result is a clear intent, so The intent of the voice request is determined to be classified as a full intent. Since the voice request has two entity values, and the semantic slots corresponding to the two entity values are both "weather location", the two entity values satisfy the same semantic slot. In the "weather" scenario, search for answers for the entity value "Beijing" and the entity value "Shanghai" respectively. The answer corresponding to the value "Shanghai" is "Thunderstorm, 21°C". Select any answer from the answers corresponding to each entity value; for the two entity values, splicing the selected two answers to obtain two sets of answers corresponding to the entity value of the voice request, and the two sets of answers are " Beijing is sunny today, 16-30 ℃; Shanghai today is thunderstorms, 21 ℃" and "Beijing tomorrow, light rain, 25 ℃; Shanghai today, thunderstorms, 21 ℃". Afterwards, a query request can be sent for the two sets of answers, for example, "Which day do you want to know the weather?", according to the user's answer of "I want to know the weather today", change the query to "Beijing is sunny today, 16-30℃; Shanghai today is Thunderstorm, 21°C" as the answer to the voice request and fed back to the user.

The voice request is "Help me check the weather in Beijing on February 1st", the semantic parsing result indicates that the scene of the voice request is "weather", and the entity values are "Beijing" and "February 1st" respectively, the intent recognition result In order to clarify the intent, it is determined that the intent of the voice request is classified as a full intent. The voice request has two entity values, the semantic slot corresponding to the entity value "Beijing" is "weather location", and the semantic slot corresponding to the entity value "February 1" is "weather time". Since the two entity values do not satisfy the same semantic slot. Therefore, in the "weather" scenario, searching for answers that satisfy both "weather location" and "weathertime", there are two answers corresponding to the entity values, namely "Beijing February 1 is sunny, 16-30℃" and " Thunderstorms in Beijing on February 1, 16-30℃”. The answer with the highest confidence is selected from the two search answers as the answer to the voice request, for example, the selected answer is "Beijing is sunny on February 1, 16-30 ℃".

Thus, the embodiments of the present invention can recognize multi-intent voice requests, improve the function of the dialogue system, and improve the intelligence of the dialogue system pair.

It should be understood that, in various embodiments of the present invention, the size of the sequence numbers of the above-mentioned processes does not mean the sequence of execution, and the execution sequence of each process should be determined by its functions and inherent logic, rather than the implementation of the present invention. The implementation of the examples constitutes no limitation.

As shown in FIG. 4, it is a schematic diagram of a question and answer processing apparatus according to an embodiment of the present invention; the apparatus 400 includes: an acquisition module 401, for acquiring a voice request; and a processing module 402, for processing the voice request, and obtaining and the voice request Corresponding intent classification; the search module 403 is configured to perform an answer search operation corresponding to the intent classification for the voice request, and feed back the search answer.

In an optional embodiment, the processing module 402 includes: a semantic parsing unit for performing semantic parsing on the voice request; an intent recognition unit for performing intent recognition on the voice request by using a model; a determining unit for converting the semantic parsing result A voice request without a scene and an intent recognition result of an ambiguous intent is determined as an incomplete intent; and a voice request with a semantic analysis result of a scene and an intent recognition result of a clear intent is determined as a complete intent.

In an optional embodiment, the search module 403 includes: a first search unit, configured to search for the answer to the voice request in the scenario indicated by the semantic analysis result if the intent classification is a complete intent; If the intent classification is an incomplete intent, predict multiple preselected scenarios corresponding to the voice request, select a preselected scenario that satisfies the first preset condition from the multiple preselected scenarios as candidate scenarios, and send a query request corresponding to the candidate scenario. Search for the answer to the voice request in the candidate scene indicated by the request result.

In an optional embodiment, the first search unit includes: an entity value unit, configured to search for an answer corresponding to the entity value in the scenario indicated by the semantic parsing result if an entity value exists in the semantic parsing result, and retrieve an answer from the searched answer Select the answer of the voice request; the scene unit is used to search for the answer corresponding to the scene indicated by the semantic parsing result if there is no entity value in the semantic parsing result, and select the answer of the voice request from the search answers.

In an optional embodiment, the entity value unit includes: a first entity value unit, used for searching for an answer corresponding to the entity value in the scenario indicated by the semantic parsing result if there is an entity value; a second entity value unit, It is used to determine whether multiple entity values satisfy the same semantic slot at the same time if there are multiple entity values; if so, search the answer for each entity value separately in the scenario indicated by the semantic parsing result, and select the corresponding answer from each entity value. Select any answer from the list; for multiple entity values, splicing the selected multiple answers to obtain the search answer corresponding to the entity value; if not, search in the scenario indicated by the semantic parsing result and satisfy multiple entity values at the same time s answer.

In an optional embodiment, the first preset condition refers to a preselected scene with the top two confidence levels and a search answer.

The above apparatus can execute the question and answer processing method provided by the embodiment of the present invention, and has functional modules and beneficial effects corresponding to executing the question and answer processing method. For technical details not described in detail in this embodiment, reference may be made to the question and answer processing method provided by the embodiment of the present invention.

As shown in FIG. 5 , which is an exemplary system architecture diagram to which the embodiments of the present invention may be applied, the system architecture 500 may include terminal devices 501 , 502 , and 503 , a network 504 , and a server 505 . The network 504 is a medium used to provide a communication link between the terminal devices 501 , 502 , 503 and the server 505 . Network 504 may include various connection types, such as wired, wireless communication links, or fiber optic cables, among others.

The user can use the terminal devices 501, 502, 503 to interact with the server 505 through the network 504 to receive or send messages and the like. Various communication client applications may be installed on the terminal devices 501 , 502 and 503 , such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only examples).

The terminal devices 501, 502, 503 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

The server 505 may be a server that provides various services, such as a background management server that supports the click events generated by the user using the terminal devices 501 , 502 , and 503 (just an example). The background management server can analyze and process the received click data, text content and other data, and feed back the processing results (such as target push information, product information—just an example) to the terminal device.

It should be noted that the question and answer processing method provided by the embodiment of the present application is generally executed by the server 505 , and accordingly, the interpretation device is generally set in the server 505 .

It should be understood that the numbers of terminal devices, networks and servers in FIG. 5 are only illustrative. There can be any number of terminal devices, networks and servers according to implementation needs.

Referring next to FIG. 6 , it shows a schematic structural diagram of a computer system suitable for implementing the terminal device or server of the embodiment. The terminal device shown in FIG. 6 is only an example, and should not impose any limitations on the functions and scope of use of the embodiments of the present invention.

As shown in FIG. 6, a computer system 600 includes a central processing unit (CPU) 601, which can be loaded into a random access memory (RAM) 603 according to a program stored in a read only memory (ROM) 602 or a program from a storage section 608 Instead, various appropriate actions and processes are performed. In the RAM 603, various programs and data necessary for the operation of the system 600 are also stored. The CPU 601 , the ROM 602 and the RAM 603 are connected to each other through a bus 604 . An input/output (I/O) interface 605 is also connected to bus 604 . The following components are connected to the I/O interface 605: an input section 606 including a keyboard, a mouse, etc.; an output section 607 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 608 including a hard disk, etc. ; and a communication section 609 including a network interface card such as a LAN card, a modem, and the like. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the I/O interface 605 as needed. A removable medium 611, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is mounted on the drive 610 as needed so that a computer program read therefrom is installed into the storage section 608 as needed.

In particular, the processes described above with reference to the flowcharts may be implemented as computer software programs in accordance with the disclosed embodiments of the present invention. For example, embodiments disclosed herein include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the method illustrated in the flowchart. In such an embodiment, the computer program may be downloaded and installed from the network via the communication portion 609 and/or installed from the removable medium 611 . When the computer program is executed by the central processing unit (CPU) 601, the above-described functions defined in the system of the present invention are performed.

It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the above two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, system or device, or a combination of any of the above. More specific examples of computer readable storage media may include, but are not limited to, electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), erasable Programmable read only memory (EPROM or flash memory), fiber optics, portable compact disk read only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing. In the present invention, a computer-readable storage medium can be any tangible medium that contains or stores a program, and the program can be used by or in conjunction with an instruction execution system, system, or device. In the present invention, however, a computer-readable signal medium may include a data signal in baseband or propagated as part of a carrier wave with computer-readable program code embodied therein. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. A computer-readable signal medium can also be any computer-readable medium, other than a computer-readable storage medium, that can transmit, propagate, or transport the program for use by or in connection with the instruction execution system, system, or device . Program code embodied on a computer readable medium may be transmitted using any suitable medium including, but not limited to, wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code that contains one or more logical functions for implementing the specified functions executable instructions. It should also be noted that, in some alternative implementations, the functions noted in the blocks may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It is also noted that each block of the block diagrams or flowchart illustrations, and combinations of blocks in the block diagrams or flowchart illustrations, can be implemented in special purpose hardware-based systems that perform the specified functions or operations, or can be implemented using A combination of dedicated hardware and computer instructions is implemented.

The modules involved in the embodiments of the present invention may be implemented in a software manner, and may also be implemented in a hardware manner. The described modules can also be provided in the processor, for example, it can be described as: a processor includes a sending module, an obtaining module, a determining module and a first processing module. Among them, the names of these modules do not constitute a limitation on the unit itself under certain circumstances. For example, the sending module can also be described as "a module that sends a request for image acquisition to the connected server".

As another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiments; or may exist alone without being assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by a device, the device includes: S101: Acquire a voice request. S102: Process the voice request to obtain an intent classification corresponding to the voice request. S103: Perform an answer search operation corresponding to the intent classification on the voice request, and feed back a search answer.

The system of the embodiment of the present invention can handle not only single-intent dialogue, but also incomplete-intent and multi-intent dialogue, thereby improving the intelligence of the dialogue system.

In the description of this specification, description with reference to the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples", etc., mean specific features described in connection with the embodiment or example , structure, material or feature is included in at least one embodiment or example of the present invention. 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, those skilled in the art may combine and combine the different embodiments or examples described in this specification, as well as the features of the different embodiments or examples, without conflicting each other.

In addition, the terms "first" and "second" are only used for descriptive purposes, and should not be construed as indicating or implying relative importance or implying the number of indicated technical features. Thus, a feature delimited with "first", "second" may expressly or implicitly include at least one of that feature. In the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited to this. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention. should be included within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (10)

1. A question-answer processing method, characterized by comprising:

acquiring a voice request;

processing the voice request to obtain an intention classification corresponding to the voice request;

and executing answer searching operation corresponding to the intention classification on the voice request, and feeding back a searching answer.

2. The method of claim 1, wherein the processing the voice request to obtain an intent classification corresponding to the voice request comprises:

performing semantic analysis on the voice request;

performing intention recognition on the voice request by utilizing a model;

determining the voice request with the semantic parsing result being scene-free and the intention recognition result being ambiguous intention as incomplete intention; and determining the voice request with the semantic parsing result of having scenes and the intention recognition result of being clear intention as a complete intention.

3. The method of claim 2, wherein performing an answer search operation on the voice request corresponding to the intent classification comprises:

if the intention classification is a complete intention, searching answers of the voice request under the scene indicated by the semantic parsing result;

if the intention classification is an incomplete intention, predicting a plurality of preselected scenes corresponding to the voice request, and selecting a preselected scene meeting a first preset condition from the preselected scenes as a candidate scene; and sending a query request corresponding to the candidate scene, and searching for an answer of the voice request under the candidate scene indicated by the query request result.

4. The method according to claim 3, wherein the searching for the answer of the voice request in the scene indicated by the semantic parsing result comprises:

if the semantic analysis result has an entity value, searching answers corresponding to the entity value in a scene indicated by the semantic analysis result, and selecting answers of the voice request from the searched answers;

and if the semantic analysis result does not have an entity value, searching answers corresponding to the scene indicated by the semantic analysis result, and selecting answers of the voice request from the searched answers.

5. The method according to claim 4, wherein if an entity value exists in the semantic analysis result, searching for an answer corresponding to the entity value in a scene indicated by the semantic analysis result comprises:

if an entity value exists, searching an answer corresponding to the entity value under the scene indicated by the semantic analysis result;

if a plurality of entity values exist, determining whether the entity values simultaneously satisfy the same semantic slot; if yes, respectively searching answers aiming at each entity value under the scene indicated by the semantic analysis result, and selecting any one answer from answers corresponding to each entity value; aiming at a plurality of entity values, splicing a plurality of selected answers to obtain search answers corresponding to the entity values; if not, searching answers meeting the entity values simultaneously under the scene indicated by the semantic analysis result.

6. The method according to claim 3, wherein the first preset condition is a preselected scene with the first two confidence levels and the search answer.

7. A question-answering processing apparatus characterized by comprising:

the acquisition module is used for acquiring a voice request;

the processing module is used for processing the voice request to obtain an intention classification corresponding to the voice request;

and the searching module is used for executing answer searching operation corresponding to the intention classification on the voice request and feeding back searching answers.

8. The apparatus of claim 7, wherein the processing module comprises:

the semantic analysis unit is used for carrying out semantic analysis on the voice request;

an intention recognition unit for performing intention recognition on the voice request by using a model;

a determining unit configured to determine a voice request, the semantic parsing result of which is scene-free and the intention recognition result of which is an ambiguous intention, as an incomplete intention; and determining the voice request with the semantic parsing result of having scenes and the intention recognition result of being clear intention as a complete intention.

9. The apparatus of claim 8, wherein the search module comprises:

the first searching unit is used for searching answers of the voice requests under the scene indicated by the semantic parsing result if the intention classification is a complete intention;

and the second searching unit is used for predicting a plurality of preselected scenes corresponding to the voice request if the intention classification is an incomplete intention, selecting the preselected scenes meeting a first preset condition from the preselected scenes as candidate scenes, sending an inquiry request corresponding to the candidate scenes, and searching answers of the voice request under the candidate scenes indicated by the inquiry request result.

10. The apparatus of claim 9, wherein the first search unit comprises:

an entity value unit, configured to search, if an entity value exists in the semantic analysis result, an answer corresponding to the entity value in a scene indicated by the semantic analysis result, and select an answer to the voice request from the search answers;

and the scene unit is used for searching answers corresponding to the scenes indicated by the semantic analysis result if the semantic analysis result does not have an entity value, and selecting the answer of the voice request from the searched answers.

Images