Information processing method and device

Abstract

The present application discloses an information processing method and device. The method includes: first, determining the first user intention according to the user input; then, querying from the pre-established user intention correspondence network whether there is a second user intention after transferring to manual service corresponding to the first user intention, and if so, determining the response information for responding to the user input according to the first user intention and the second user intention; then, outputting the response information. Among them, each relationship in the user intention correspondence network is the correspondence between the user intention before transferring to manual service and the user intention after transferring to manual service. Since the second user intention is the user intention after further determination after transferring to manual service, it is more likely to reflect the user's true user intention. As a beneficial supplement to the first user intention, jointly determining the response information can reduce the probability of users transferring to manual service, thereby greatly reducing labor costs and improving the utilization rate of intelligent customer service.

Description

Technical Field

[0001] The present application relates to computer information processing technology, and in particular to an information processing method and device. Background Art

[0002] With the development and popularization of artificial intelligence technology, intelligent dialogue systems, especially intelligent customer service, are increasingly widely used in various fields. However, due to the different language expression patterns of each person, especially in more casual dialogue scenarios such as spoken language, it is very likely that due to various reasons such as missing grammatical structure, overly colloquial language, and sudden jumps in context, intelligent customer service may not be able to correctly understand the user's intention in many cases, and has to switch to manual service to continue answering the user's questions.

[0003] When intelligent customer service is converted to manual service too many times, on the one hand, labor costs will increase and the expected benefits of using intelligent customer service will not be achieved; on the other hand, after transferring to manual service, users are often required to provide relevant information again, which takes double the time, resulting in a significant drop in user satisfaction. Summary of the invention

[0004] The inventors of the present applicant creatively provide an information processing method and device.

[0005] According to a first aspect of an embodiment of the present application, there is provided an information processing method, the method comprising: receiving user input; determining a first user intention of the user input; determining whether there is a second user intention corresponding to the first user intention based on a user intention correspondence network, and if so, determining response information for responding to the user input based on the first user intention and the second user intention, the user intention correspondence in the user intention correspondence network comprising a correspondence between the user intention before transferring to manual service and the user intention after transferring to manual service; and outputting the response information.

[0006] According to an embodiment of the present application, before determining whether there is a second user intent corresponding to the first user intent based on the user intention correspondence network, the method also includes: obtaining a third user intent before transferring to manual service based on the user's historical conversation records; obtaining a fourth user intent after transferring to manual service based on the user's historical conversation records; and establishing a correspondence between the third user intent and the fourth user intent.

[0007] According to an embodiment of the present application, the user intention before transferring to manual service is recorded in the user history conversation record. Accordingly, based on the user history conversation record, the third user intention before transferring to manual service is obtained, including: obtaining the user intention before transferring to manual service from the user history conversation record as the third user intention.

[0008] According to an embodiment of the present application, obtaining a fourth user intent after transfer to manual service based on the user's historical conversation record includes: obtaining the fourth user intent after transfer to manual service based on the user's historical conversation record input by the manual service personnel.

[0009] According to an embodiment of the present application, obtaining a fourth user intention after transferring to manual service based on the user's historical conversation records includes: obtaining the fourth user intention after transferring to manual service based on the conversation and language model between the user and the manual service personnel in the user's historical conversation records.

[0010] According to one embodiment of the present application, the second user intent includes at least two user intentions, and accordingly, response information for responding to user input is determined based on the first user intent and the second user intent, including: obtaining user information and / or context information of the current conversation; filtering the second user intent based on the user information and / or the context information of the current conversation to obtain a fifth user intent; and determining response information for responding to user input based on the first user intent and the fifth user intent.

[0011] According to an embodiment of the present application, the fifth user intent includes at least two user intentions, and accordingly, response information for responding to user input is determined according to the first user intent and the fifth user intent, including: sorting the user intents in the fifth user intent according to the degree of match between each user intent in the fifth user intent and the first user intent; determining response information for responding to user input according to the first user intent and the top N user intentions, where N is a natural number greater than or equal to 1.

[0012] According to an embodiment of the present application, the matching degree includes the number of matches with the first user's intention in the historical conversation records.

[0013] According to an embodiment of the present application, the method further includes: determining whether there is a second user intent corresponding to the first user intent based on a user intent correspondence network; if not, determining response information for responding to the user input based on the first user intent.

[0014] According to a second aspect of an embodiment of the present application, an information processing device is provided, which includes: a receiving module for receiving user input; a first user intention determination module for determining the first user intention of the user input; a response information determination module for determining whether there is a second user intention corresponding to the first user intention based on a user intention correspondence network, and if so, determining response information for responding to the user input based on the first user intention and the second user intention, the user intention correspondence in the user intention correspondence network including the correspondence between the user intention before transferring to manual service and the user intention after transferring to manual service; and an output module for outputting response information.

[0015] The embodiment of the present application provides an information processing method and device, the method comprising: first, determining a first user intention according to a user input; then, querying from a pre-established user intention correspondence network whether there is a second user intention after switching to manual service corresponding to the first user intention, and if so, determining response information for responding to the user input according to the first user intention and the second user intention; then, outputting the response information. Each relationship in the user intention correspondence network is a correspondence between a user intention before switching to manual service and a user intention after switching to manual service.

[0016] Since the second user intention is the user intention that has been further determined after being transferred to manual service, it is more likely to reflect the user's true user intention. As a beneficial supplement or correction to the first user intention, jointly determining the response information can increase the possibility of being close to user wishes and solving user problems, thereby reducing the probability of users transferring to manual service, greatly reducing labor costs, improving the utilization rate of intelligent customer service, and further improving user satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present application will become readily understood. In the accompanying drawings, several embodiments of the present application are shown in an exemplary and non-limiting manner, wherein:

[0018] In the drawings, the same or corresponding reference numerals represent the same or corresponding parts.

[0019] Figure 1 This is a schematic diagram of the implementation process of an embodiment of the information processing method of this application;

[0020] Figure 2 This is a schematic diagram of the implementation flow of another embodiment of the information processing method of the present application;

[0021] Figure 3 This is a schematic diagram of the structure of an information processing device according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to make the purpose, features, and advantages of the present application more obvious and easy to understand, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0023] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0024] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0025] Figure 1 The implementation process of an embodiment of the information processing method of the present application is shown. Figure 1 The method includes: operation 110, receiving user input; operation 120, determining a first user intention of the user input; operation 130, determining whether there is a second user intention corresponding to the first user intention according to a user intention correspondence network, and if so, determining response information for responding to the user input according to the first user intention and the second user intention, the user intention correspondence in the user intention correspondence network including the correspondence between the user intention before transferring to manual service and the user intention after transferring to manual service; operation 140, outputting the response information.

[0026] In operation 110, the user input mainly refers to the content input by the user in the current conversation, which may be text, voice, image, or one or several options determined from multiple options.

[0027] Usually, the user input corresponds to one or more user intentions, which represent the key information, true meaning or purpose that the user input wants to express. In the process of human-computer dialogue, the recognition of user intention is very critical. Whether the user's true user intention can be accurately recognized directly determines whether the information that the user really wants can be given or whether the subsequent dialogue can proceed smoothly.

[0028] Operation 120 is used to determine the user intention corresponding to the user input, that is, the first user intention, according to the user input received in operation 110. The first user intention may be one user intention or multiple user intentions.

[0029] Operation 120 is a relatively basic operation in current human-computer dialogue systems, and any applicable technology may be used, such as user intention recognition based on multiple classifications or user intention recognition based on machine learning.

[0030] The user intent determined by user intent recognition may have limited accuracy due to deviations caused by various factors such as classification rules, classification methods or training data. If the accuracy of user intent recognition is improved simply by improving classification rules, classification methods or training data, it is usually complicated and has little effect.

[0031] Therefore, the inventor of the present application creatively thought that it is possible to use existing conversation log information or other methods to obtain historical conversation records, and obtain the user intention before switching to manual service and the user intention after switching to manual service from the historical conversation records. Then, based on the historical conversation records, the internal connection or law between the user intentions before and after switching to manual service is analyzed and learned, and a corresponding relationship is formed between the user intention before switching to manual service and the user intention after switching to manual service. In this way, more possible user intentions can be found based on the first user intention as a supplement or correction to the first user intention, thereby increasing the probability of accurately identifying the user intention, thereby improving the accuracy of user intention identification and user experience.

[0032] In operation 130, the user intention correspondence network is a correspondence network of user intentions before switching to manual service and user intentions after switching to manual service pre-established according to the above-mentioned inventive concept. By performing relationship calculation based on the correspondence network, it can be determined whether there is a user intention after switching to manual service (i.e., the second user intention) corresponding to the first user intention.

[0033] For example, assuming that the corresponding relationship network is constructed in the form of a graph and stored in a relational database, it is possible to determine whether there is a point in the graph representing the first user intention and a point connected to or reachable from the point by querying the relational database to determine whether there is a second user intention corresponding to the first user intention.

[0034] If so, response information for responding to the user input may be determined according to the first user intention and the second user intention, and the response information may be output through operation 140. The response information for responding to the user input may be a statement that directly answers the user's question, or a statement that asks for more information to further advance the conversation, or a statement that provides multiple possible options for the user to select or confirm.

[0035] Since the second user intention is the user intention after being transferred to manual service and further determined, it is more likely to reflect the user's real user intention. As a beneficial supplement to the first user intention, jointly determining the response information can increase the possibility of being close to the user's wishes and solving user problems, thereby reducing the probability of users transferring to manual service, greatly reducing labor costs, increasing the utilization rate of intelligent customer service, and further improving user satisfaction.

[0036] It should be noted that Figure 1 The embodiment shown is only the most basic embodiment of the information processing method of the present application, and implementers can further refine and expand it on this basis.

[0037] According to an embodiment of the present application, before determining whether there is a second user intent corresponding to the first user intent based on the user intention correspondence network, the method also includes: obtaining a third user intent before transferring to manual service based on the user's historical conversation records; obtaining a fourth user intent after transferring to manual service based on the user's historical conversation records; and establishing a correspondence between the third user intent and the fourth user intent.

[0038] By obtaining the user intentions before and after manual service from historical conversation records, and then analyzing and learning them, we can get the correspondence between the user intentions before manual service and the user intentions after manual service.

[0039] For example, after the user enters the intelligent customer service, the input content is "My mobile phone volume is low", and the intelligent customer service understands the user's intention as "Call volume problem", and then gives the corresponding answer. At this time, the user transfers to manual operation, and the user intention obtained from the manual customer service is "Mobile phone loudspeaker problem". After this type of data is accumulated for a period of time, it can be considered that "Call volume problem" and "Mobile phone loudspeaker problem" have a certain correlation. At this time, the corresponding relationship between the user intention before manual service and the user intention after manual service can be established.

[0040] In addition, based on the correspondence between user intentions before and after the transfer to manual service, other information with obvious distinctions such as country and language can also be considered; then, based on the above information, a user intention correspondence network can be constructed. The user intention correspondence network can be constructed through a graph or any other applicable method, and stored in a data storage system such as a database, LDAP server, network disk, etc. for subsequent manual dialogue.

[0041] After the user intention correspondence network is created, it can be continuously updated, supplemented and improved based on the continuously generated conversation records; then, it can be further applied to actual human-computer conversations, thus forming an information closed loop of learning and continuous optimization, which continuously improves the accuracy of user intention recognition.

[0042] According to an embodiment of the present application, the user intention before transferring to manual service is recorded in the user history conversation record. Accordingly, based on the user history conversation record, the third user intention before transferring to manual service is obtained, including: obtaining the user intention before transferring to manual service from the user history conversation record as the third user intention.

[0043] Usually, during the human-computer dialogue process, user input and user intent determined based on the user input are recorded to form a user history dialogue record. In this case, the record of transfer to manual service and the user intent before transfer to manual service can be directly located from the user history dialogue record. When obtaining the user intent before transfer to manual service, the last user intent before transfer to manual service can be obtained, or the last few user intents before transfer to manual service can be obtained. The implementer can flexibly determine it according to the specific implementation requirements, implementation conditions and implementation effects.

[0044] According to an embodiment of the present application, obtaining a fourth user intent after transfer to manual service based on the user's historical conversation record includes: obtaining the fourth user intent after transfer to manual service based on the user's historical conversation record input by the manual service personnel.

[0045] In the embodiment of the present application, after the transfer to manual service, the manual service personnel will continue to provide service to the user, and after the service is completed, the manual service personnel will manually input or confirm one or more user intentions from a number of preset user intentions as the user intention after the transfer to manual service, and store them in the user's historical conversation record. In this way, the position of the transfer to manual service can be located from the user's historical conversation record, and the user intention entered by the manual service personnel after the transfer to manual service can be obtained as the user intention after the transfer to manual service.

[0046] Although this implementation method will increase manual operations, the user intent input by manual service personnel will be closer to the user's actual user intent and have a higher accuracy rate.

[0047] According to an embodiment of the present application, obtaining a fourth user intention after transferring to manual service based on the user's historical conversation records includes: obtaining the fourth user intention after transferring to manual service based on the conversation and language model between the user and the manual service personnel in the user's historical conversation records.

[0048] In this embodiment, the human service personnel are not required to manually input the user intention after the transfer to human service, but the conversation between the user and the human service personnel will be saved in the user's historical conversation record. In this way, the location of the transfer to human service can be located, the conversation with the human service personnel after the location is obtained, and the fourth user intention after the transfer to human service can be obtained by semantic analysis using the language model.

[0049] The language model used here can be the language model used to identify user intentions in the current human-computer dialogue system. In this way, there is no need to introduce more models, avoiding more model maintenance costs; the language model can also be trained using the dialogue records between users and human service personnel to further improve the accuracy of the language model and apply it to the actual human-computer dialogue process.

[0050] According to one embodiment of the present application, the second user intent includes at least two user intentions, and accordingly, response information for responding to user input is determined based on the first user intent and the second user intent, including: obtaining user information and / or context information of the current conversation; filtering the second user intent based on the user information and / or the context information of the current conversation to obtain a fifth user intent; and determining response information for responding to user input based on the first user intent and the fifth user intent.

[0051] When the user's historical conversation record is long, there may be multiple second user intentions corresponding to the first user intention. At this time, it may be necessary to filter the multiple second user intentions to filter out the user intentions that are obviously not applicable to the current user or are not suitable for the current conversation scene or conversation context information.

[0052] The user information mainly refers to the user's geographical location, the language used by the user, the user's gender or the preferences set by the user, etc., which may affect the user's intended selection. For example, when a user in Beijing wants to obtain the after-sales service center of a certain product, the after-sales service center located in Beijing and closer to the user's current location can be output preferentially.

[0053] The contextual information of the current conversation can also further help determine and filter the queried user intent.

[0054] Filtering the acquired second user intention can make the user intention involved in subsequent calculations more accurate, saving a lot of computing resources and greatly shortening the computing time and improving the response speed without affecting the accuracy.

[0055] According to an embodiment of the present application, the fifth user intent includes at least two user intentions, and accordingly, response information for responding to user input is determined according to the first user intent and the fifth user intent, including: sorting the user intents in the fifth user intent according to the degree of match between each user intent in the fifth user intent and the first user intent; determining response information for responding to user input according to the first user intent and the top N user intentions, where N is a natural number greater than or equal to 1.

[0056] There may still be many fifth user intentions obtained after filtering. Therefore, the fifth user intentions may be sorted according to the matching degree between each user intention and the first user intention, and the user intentions with a greater success probability may be used first.

[0057] Among them, the degree of match with the first user intention mainly refers to the historical probability of becoming the user intention after transferring to manual service corresponding to the first user intention; or predicting the possibility of becoming the user intention that the user really wants to express if the user transfers to manual service in the current dialogue scenario based on semantics and the current dialogue context.

[0058] According to an embodiment of the present application, the matching degree includes the number of matches with the first user's intention in the historical conversation records.

[0059] The number of matches with the first user's intention mainly refers to the number of user intentions after being transferred to manual service in the historical conversation records that correspond to the first user's intention.

[0060] For example, suppose that the user intention before and after each transfer to manual service in the historical conversation record is stored as a record such as <time of transfer to manual service, user intention before transfer to manual service, user intention after transfer to manual service>, query the records in which the user intention before transfer to manual service is the first user intention, and classify and count the records according to "user intention after transfer to manual service" to obtain the number of records of each user intention after transfer to manual service. This number can be used as the number of matches between each user intention after transfer to manual service and the first user intention.

[0061] Theoretically, the more matches there are with the first user intent in the history records, the more likely it is that a real user intent is often mistaken for the first user intent. In this way, it can be inferred that the user intent may also be the real user intent of the current conversation. Then, the user intent and the first user intent can be used together as possible user intents, thereby further increasing the probability of finding the real user intent of the user.

[0062] According to an embodiment of the present application, the method further includes: determining whether there is a second user intent corresponding to the first user intent based on a user intent correspondence network; if not, determining response information for responding to the user input based on the first user intent.

[0063] When there are not many historical conversation records, or the first user intention is very clear and is less likely to be misunderstood, it is very likely that there is no second user intention corresponding to the first user intention. In this case, the response information for responding to the user input can be determined based on the original processing logic only according to the first user intention.

[0064] For example, the user input is "I want to make a free call", and the user intent understood by the intelligent customer service is "make a free call". No user intent related to "make a free call" is found in the user intent correspondence network. At this time, only the response information needs to be determined based on the user intent of "make a free call".

[0065] The above embodiments are for how to Figure 1 The exemplary descriptions of how to further refine and expand the basic embodiments shown are provided. Implementers may also combine various implementation methods in the above embodiments to form new embodiments according to specific implementation conditions and needs to achieve more ideal implementation effects.

[0066] Figure 2 Another embodiment of the information processing method of the present application is shown. This embodiment combines the above multiple implementation methods to finally form an embodiment with better implementation effect.

[0067] In this embodiment, the human-computer dialogue system will store the dialogue between the user and the intelligent customer service in the dialogue log. Before the dialogue with the user, a collection of user intentions before and after the transfer to manual service and a corresponding relationship network of user intentions before and after the transfer to manual service have been established based on the user's historical dialogue log, wherein the collection of user intentions before and after the transfer to manual service only includes the user intentions before and after the transfer to manual service, which can be regarded as the primary key directory or index table of the corresponding relationship network of user intentions before and after the transfer to manual service. In this way, querying the table can quickly determine whether the user intention determined by the current dialogue has a corresponding user intention after the transfer to manual service, without having to perform complex relationship calculations based on the corresponding relationship network of user intentions before and after the transfer to manual service.

[0068] In this embodiment, the collection of user intentions before and after manual service and the corresponding relationship network of user intentions before and after manual service are stored in a relational database. The collection of user intentions before and after manual service is a table in the relational database, and the corresponding relationship network of user intentions before and after manual service is constructed in a graph manner. The corresponding relationship between the user intention before manual service and the user intention after manual service is realized in the primary and foreign key manner in the database, and the corresponding relationship calculation is obtained by querying the database.

[0069] like Figure 2 As shown, for the currently ongoing session process, the processing flow of this embodiment mainly includes:

[0070] Operation 210, reading the current conversation log (for obtaining received user input);

[0071] Operation 220, determining a user intention (first user intention) of the current conversation;

[0072] Operation 230, querying the collection of user intentions before and after the transfer to manual service;

[0073] Operation 240, calculating the user intention after transferring to manual service corresponding to the user intention of the current conversation according to the user intention relationship network before and after transferring to manual service;

[0074] Operation 250, filtering the user intention after being transferred to manual service;

[0075] Operation 260, sorting the user intentions after being transferred to manual service;

[0076] Operation 270 , calculating and executing an interaction strategy according to the user intention of the current conversation and the user intention after the final determination of the transfer to manual service.

[0077] It should be noted that Figure 2 The applications shown are only exemplary descriptions of the information processing method of the present application and are not intended to limit the implementation methods and application scenarios of the information processing method of the present application. The implementer may adopt any applicable implementation method and apply it to any applicable application scenario according to specific implementation conditions.

[0078] Furthermore, the present application also provides an information processing device, such as Figure 3 As shown, the device 30 includes: a receiving module 301, receiving user input; a first user intention determination module 302, used to determine the first user intention of the user input; a response information determination module 303, used to determine whether there is a second user intention corresponding to the first user intention based on a user intention correspondence network, if so, then determine the response information for responding to the user input based on the first user intention and the second user intention, the user intention correspondence in the user intention correspondence network includes the correspondence between the user intention before transferring to manual service and the user intention after transferring to manual service; an output module 304, used to output the response information.

[0079] According to one embodiment of the present application, the device 30 also includes: a third user intention acquisition module, which is used to acquire the third user intention before transferring to manual service based on the user's historical conversation records; a fourth user intention acquisition module, which is used to acquire the fourth user intention after transferring to manual service based on the user's historical conversation records; and a correspondence establishment module, which is used to establish a correspondence between the third user intention and the fourth user intention.

[0080] According to an embodiment of the present application, the user's intention before transferring to manual service is recorded in the user's historical conversation record. Accordingly, the third user intention acquisition module is specifically used to acquire the user's intention before transferring to manual service from the user's historical conversation record as the third user intention.

[0081] According to an embodiment of the present application, the fourth user intention acquisition module is specifically used to acquire the fourth user intention after transfer to manual service based on the user intention input by the manual service personnel in the user's historical conversation record.

[0082] According to an embodiment of the present application, the fourth user intention acquisition module is specifically used to acquire the fourth user intention after transfer to manual service based on the conversation and language model between the user and the manual service personnel in the user's historical conversation record.

[0083] According to one embodiment of the present application, the response information determination module 303 includes: an additional information acquisition submodule, used to acquire user information and / or context information of the current conversation; a user intention filtering submodule, used to filter the second user intention according to the user information and / or the context information of the current conversation to obtain the fifth user intention; a response information determination submodule, used to determine the response information for responding to the user input according to the first user intention and the fifth user intention.

[0084] According to one embodiment of the present application, the response information determination module 303 includes: a user intention sorting submodule, used to sort the user intentions in the fifth user intentions according to the matching degree of each user intention in the fifth user intentions with the first user intention; a response information determination submodule, used to determine the response information used to respond to the user input according to the first user intention and the top N user intentions, where N is a natural number greater than or equal to 1.

[0085] According to one embodiment of the present application, the response information determination module 303 is also used to determine whether there is a second user intent corresponding to the first user intent based on the user intent correspondence network. If not, the response information used to respond to the user input is determined based on the first user intent.

[0086] According to a third aspect of an embodiment of the present application, a computer-readable storage medium is provided, the storage medium comprising a set of computer-executable instructions, which, when executed, are used to execute any of the above-mentioned information processing methods applied to a recommendation system.

[0087] It should be pointed out here that the above description of the information processing device embodiment and the above description of the computer-readable storage medium embodiment are similar to the description of the aforementioned method embodiment and have similar beneficial effects as the aforementioned method embodiment, so they are not described in detail. For the technical details that have not been disclosed in the description of the information processing device embodiment and the description of the computer-readable storage medium embodiment of this application, please refer to the description of the aforementioned method embodiment of this application for understanding, and in order to save space, they are not described in detail.

[0088] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0089] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another device, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0090] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0091] In addition, all functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may be a separate unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0092] A person skilled in the art can understand that all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: mobile storage medium, read-only memory (ROM), disk or optical disk, etc. Various media that can store program codes.

[0093] Alternatively, if the above-mentioned integrated unit of the present application is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present application can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods of each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage media, ROM, magnetic disks or optical disks.

[0094] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An information processing method, the method comprising: Receive user input; determining a first user intention of the user input; Determine whether there is a second user intent corresponding to the first user intent according to a user intent correspondence network, and if so, determine response information for responding to the user input according to the first user intent and the second user intent, wherein the user intent correspondence in the user intent correspondence network includes a correspondence between the user intent before transferring to manual service and the user intent after transferring to manual service; The response information is output.

2. The method according to claim 1, before determining whether there is a second user intent corresponding to the first user intent according to the user intent correspondence network, the method further comprises: Based on the user's historical conversation records, obtain the third-party user's intention before transferring to manual service; According to the user's historical conversation records, obtain the fourth user's intention after transferring to manual service; A corresponding relationship between the third user intention and the fourth user intention is established.

3. The method according to claim 2, wherein the user's historical conversation record records the user's intention before transferring to manual service. Accordingly, the third user's intention before transferring to manual service is obtained based on the user's historical conversation records, including: The user intention before transferring to manual service is obtained from the user historical conversation record as the third user intention.

4. The method according to claim 2, wherein obtaining the fourth user intention after transferring to manual service according to the user's historical conversation record comprises: According to the user intent input by the manual service personnel in the user's historical conversation record, the fourth user intent after transfer to manual service is obtained.

5. The method according to claim 2, wherein obtaining the fourth user intention after transferring to manual service according to the user's historical conversation record comprises: According to the conversation between the user and the manual service personnel and the language model in the user's historical conversation records, the fourth user intention after transferring to the manual service is obtained.

6. The method according to claim 1, wherein the second user intent includes at least two user intents, and accordingly, determining the response information for responding to the user input according to the first user intent and the second user intent comprises: Get user information and / or context information of the current conversation; Filtering the second user intent according to the user information and / or context information of the current conversation to obtain a fifth user intent; Response information for responding to the user input is determined according to the first user intention and the fifth user intention.

7. The method according to claim 6, wherein the fifth user intent includes at least two user intents, and accordingly, determining the response information for responding to the user input according to the first user intent and the fifth user intent comprises: sorting the user intents in the fifth user intents according to a matching degree between each user intent in the fifth user intent and the first user intent; Response information for responding to the user input is determined according to the first user intent and the top N user intents, where N is a natural number greater than or equal to 1.

8. According to the method of claim 7, the matching degree includes the number of matches with the first user intention in the historical conversation records.

9. The method according to claim 1, further comprising: Determine whether there is a second user intent corresponding to the first user intent according to the user intent correspondence network; if not, determine response information for responding to the user input according to the first user intent.

10. An information processing device, comprising: A receiving module, used for receiving user input; A first user intention determination module, used to determine the first user intention input by the user; a response information determination module, configured to determine whether there is a second user intention corresponding to the first user intention according to a user intention correspondence network, and if so, determine response information for responding to the user input according to the first user intention and the second user intention, wherein the user intention correspondence in the user intention correspondence network includes a correspondence between the user intention before transferring to manual service and the user intention after transferring to manual service; An output module is used to output the response information.

Images