Financial service application intention recognition method and device, equipment and storage medium

Abstract

The invention relates to an intention recognition method and device for a financial service application, computer equipment, a storage medium and a computer program product, relates to the field of artificial intelligence, and can improve the intention recognition efficiency. The method comprises the steps of obtaining a to-be-recognized text which is input by a user in a financial service application and is to be subjected to intention recognition; comparing the text feature corresponding to the to-be-recognized text with text features corresponding to a plurality of financial industry knowledge texts in a pre-constructed financial industry knowledge base, and determining a target financial industry knowledge text associated with the to-be-recognized text according to the similarity of the text features obtained by comparison; inputting the target financial industry knowledge text and the to-be-recognized text into a pre-trained universal language model, and determining context semantics of the to-be-recognized text by the universal language model according to the target financial industry knowledge text, and determining a user intention corresponding to the to-be-recognized text according to the context semantics.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method, apparatus, computer equipment, storage medium and computer program product for intent recognition in financial services applications. Background Art

[0002] With the development of computer technology, in order to provide users with corresponding search results or reply content more efficiently, intent recognition can be performed by training natural language processing models.

[0003] In related technologies, in order to obtain better recognition results for information entered by users in financial service applications, it is often necessary to obtain a large amount of financial field training data and manually set training labels. After multiple training sessions, intent recognition is performed through the trained model.

[0004] However, this method has high training costs and is difficult to efficiently identify the intent of information entered by users in financial service applications. Summary of the invention

[0005] Based on this, it is necessary to provide an intent recognition method, apparatus, computer device, computer-readable storage medium and computer program product for financial service applications that can improve the efficiency of intent recognition in response to the above-mentioned technical problems.

[0006] In a first aspect, the present application provides a method for identifying intent of a financial service application. The method comprises:

[0007] Obtaining text to be recognized that is input by a user in a financial service application and to be recognized by intent;

[0008] Comparing the text features corresponding to the text to be recognized with the text features corresponding to multiple financial industry knowledge texts in a pre-built financial industry knowledge base, and determining the target financial industry knowledge text associated with the text to be recognized based on the text feature similarity obtained by the comparison;

[0009] The target financial industry knowledge text and the text to be recognized are input into a pre-trained general language model, and the general language model determines the contextual semantics of the text to be recognized based on the target financial industry knowledge text, and determines the user intention corresponding to the text to be recognized based on the contextual semantics.

[0010] In one embodiment, comparing the text features corresponding to the text to be recognized with the text features corresponding to a plurality of financial industry knowledge texts in a pre-built financial industry knowledge base includes:

[0011] The text features corresponding to the text to be recognized are compared with the text features corresponding to each financial industry knowledge text in the pre-built financial industry knowledge base to obtain the text feature similarity between the text to be recognized and each financial industry knowledge text.

[0012] In one embodiment, comparing the text features corresponding to the text to be recognized with the text features corresponding to a plurality of financial industry knowledge texts in a pre-built financial industry knowledge base includes:

[0013] Determine the knowledge labels corresponding to each category of financial industry knowledge in the pre-built financial industry knowledge base;

[0014] Matching the text to be recognized with each knowledge tag, and determining the target category financial industry knowledge corresponding to the matched knowledge tag;

[0015] The text features corresponding to the text to be recognized are compared with the text features corresponding to the financial industry knowledge text of the target category financial industry knowledge.

[0016] In one embodiment, the financial industry knowledge base is constructed by the following steps:

[0017] Clustering multiple pre-acquired financial industry knowledge texts to obtain multiple categories of financial industry knowledge and financial industry knowledge texts corresponding to each category of financial industry knowledge;

[0018] For each category of financial industry knowledge, a text vector corresponding to each financial industry knowledge text of the category of financial industry knowledge is obtained, and based on multiple text vectors, a text feature corresponding to the financial industry knowledge text of the category of financial industry knowledge is determined.

[0019] In one embodiment, determining the target financial industry knowledge text associated with the to-be-recognized text according to the text feature similarity obtained by comparison includes:

[0020] Determine the difficulty of intention recognition of the text to be recognized, and determine the number K of text screening corresponding to the difficulty of intention recognition; K is a positive integer, and K is positively correlated with the difficulty of intention recognition;

[0021] According to the text feature similarity obtained by comparison, K financial industry knowledge texts with the highest text feature similarity are obtained from the multiple financial industry knowledge texts to obtain the financial industry knowledge text associated with the text to be identified.

[0022] In one embodiment, obtaining the text to be recognized that is input by the user in the financial service application and to be recognized by the intent includes:

[0023] The text information entered by the user in the search box of the financial service application is used as the text to be recognized for intent recognition;

[0024] and / or,

[0025] The text information input by the user in the dialog box of the financial service application is used as the text to be recognized for intent recognition.

[0026] In a second aspect, the present application also provides an intention recognition device for a financial service application. The device comprises:

[0027] A text acquisition module, used to acquire text to be recognized that is input by a user in a financial service application and is to be recognized by intent;

[0028] A feature comparison module is used to compare the text features corresponding to the text to be identified with the text features corresponding to multiple financial industry knowledge texts in a pre-built financial industry knowledge base, and determine the target financial industry knowledge text associated with the text to be identified based on the text feature similarity obtained by the comparison;

[0029] The intention recognition module is used to input the target financial industry knowledge text and the text to be recognized into a pre-trained general language model, and the general language model determines the contextual semantics of the text to be recognized based on the target financial industry knowledge text, and determines the user intention corresponding to the text to be recognized based on the contextual semantics.

[0030] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0031] Obtaining text to be recognized that is input by a user in a financial service application and to be recognized by intent;

[0032] Comparing the text features corresponding to the text to be recognized with the text features corresponding to multiple financial industry knowledge texts in a pre-built financial industry knowledge base, and determining the target financial industry knowledge text associated with the text to be recognized based on the text feature similarity obtained by the comparison;

[0033] The target financial industry knowledge text and the text to be recognized are input into a pre-trained general language model, and the general language model determines the contextual semantics of the text to be recognized based on the target financial industry knowledge text, and determines the user intention corresponding to the text to be recognized based on the contextual semantics.

[0034] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0035] Obtaining text to be recognized that is input by a user in a financial service application and to be recognized by intent;

[0036] Comparing the text features corresponding to the text to be recognized with the text features corresponding to multiple financial industry knowledge texts in a pre-built financial industry knowledge base, and determining the target financial industry knowledge text associated with the text to be recognized based on the text feature similarity obtained by the comparison;

[0037] The target financial industry knowledge text and the text to be recognized are input into a pre-trained general language model, and the general language model determines the contextual semantics of the text to be recognized based on the target financial industry knowledge text, and determines the user intention corresponding to the text to be recognized based on the contextual semantics.

[0038] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0039] Obtaining text to be recognized that is input by a user in a financial service application and to be recognized by intent;

[0040] Comparing the text features corresponding to the text to be recognized with the text features corresponding to multiple financial industry knowledge texts in a pre-built financial industry knowledge base, and determining the target financial industry knowledge text associated with the text to be recognized based on the text feature similarity obtained by the comparison;

[0041] The target financial industry knowledge text and the text to be recognized are input into a pre-trained general language model, and the general language model determines the contextual semantics of the text to be recognized based on the target financial industry knowledge text, and determines the user intention corresponding to the text to be recognized based on the contextual semantics.

[0042] The above-mentioned intent recognition method, device, computer equipment, storage medium and computer program product of the financial service application can obtain the text to be recognized that is input by the user in the financial service application for intent recognition, and then compare the text features corresponding to the text to be recognized with the text features corresponding to multiple financial industry knowledge texts in the pre-constructed financial industry knowledge base, and determine the target financial industry knowledge text associated with the text to be recognized based on the text feature similarity obtained by the comparison. After that, the target financial industry knowledge text and the text to be recognized can be input into a pre-trained general language model, and the general language model determines the contextual semantics of the text to be recognized based on the target financial industry knowledge text, and determines the user intent corresponding to the text to be recognized based on the contextual semantics. In the present application, by inputting the target financial industry knowledge text that has a language association with the text to be recognized into the general language model, it is possible to adapt the language model to the financial field and understand the target financial industry knowledge without retraining the general language model, thereby helping to quickly understand the user intent of the user who inputs the text to be recognized in the financial service application and improve the efficiency of intent recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A flowchart of a method for identifying intent of a financial service application in one embodiment;

[0044] Figure 2 A flowchart of a text feature comparison step in one embodiment;

[0045] Figure 3 A schematic diagram of a process of building a financial industry knowledge base in one embodiment;

[0046] Figure 4 A structural block diagram of an intention recognition device for a financial service application in one embodiment;

[0047] Figure 5 is an internal structure diagram of a computer device in one embodiment;

[0048] Figure 6 The figure is a diagram of the internal structure of another computer device in one embodiment. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0050] In one embodiment, Figure 1As shown, a method for identifying intent of a financial service application is provided. This embodiment uses the method applied to a terminal as an example. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0051] S101, obtaining text to be recognized that is input by a user in a financial service application and for which intent recognition is to be performed.

[0052] In some embodiments, a user may input text in a financial service application, and when it is detected that the user has completed the triggering operation of the current text input, the text input by the user may be used as the text to be recognized for intent recognition, for example, when it is detected that the user clicks the "Send" button or the "Search" button in the financial service application, it may be determined that the triggering operation of the text input completion is detected. It is understandable that the way in which the user inputs text in the financial service application may include at least one of the following: text input, voice input.

[0053] In some embodiments, the text input by the user can be preprocessed and the preprocessed text can be used as the text to be recognized. For example, the word order and typos can be adjusted on the text input by the user and the processed text can be used as the text to be recognized.

[0054] S102, comparing text features corresponding to the text to be recognized with text features corresponding to multiple financial industry knowledge texts in a pre-built financial industry knowledge base, and determining a target financial industry knowledge text associated with the text to be recognized based on the text feature similarity obtained by the comparison.

[0055] After obtaining the text to be recognized, the text features corresponding to the text to be recognized can be obtained. In some embodiments, the semantic representation vector corresponding to the text to be recognized can be obtained through an existing text vector determination model, and the vector can be used as the text feature corresponding to the text to be recognized. For example, the embedded representation of the text to be recognized can be determined, and the embedded representation can be used as the text feature corresponding to the text to be recognized.

[0056] Then, the text features of the text to be identified can be compared with the text features corresponding to multiple financial industry knowledge texts in the pre-constructed financial industry knowledge base. Specifically, multiple financial industry knowledge texts can be obtained in advance. In some optional embodiments, the financial industry knowledge texts can be obtained based on financial industry textbooks, financial industry books, and financial industry multimedia information. For example, the financial industry video can be processed by semantic conversion to text or image-text recognition to obtain the corresponding financial industry knowledge text. For the acquired financial industry knowledge text, the corresponding text features can be determined, and thus a financial industry knowledge base can be constructed based on multiple financial industry knowledge texts and their corresponding text features. Among them, the text features corresponding to the financial industry knowledge text can be obtained by using the existing text vector determination model to obtain the corresponding semantic representation vector, and the vector is used as the text feature corresponding to the financial industry knowledge text; for example, the embedded representation of the financial industry knowledge text can be determined, and the embedded representation can be used as the text feature corresponding to the financial industry knowledge text.

[0057] Then, based on the comparison results, the text feature similarity between the text features of the text to be identified and the text features corresponding to each financial industry knowledge text can be obtained. Based on multiple text feature similarities, the financial industry knowledge text associated with the text to be identified can be obtained from multiple financial industry knowledge texts as the target financial industry knowledge text.

[0058] S103, input the target financial industry knowledge text and the text to be recognized into a pre-trained general language model, and the general language model determines the contextual semantics of the text to be recognized based on the target financial industry knowledge text, and determines the user intention corresponding to the text to be recognized based on the contextual semantics.

[0059] In this step, a general language model can be obtained in advance. The general language model can be understood as a model obtained through training based on a large amount of general corpus. The general language model can perform basic and conventional language processing tasks, but for language processing tasks in vertical fields, due to the lack of training data in relevant vertical fields, its task processing effect is general. In related technologies, for vertical fields, such as intent recognition tasks in the financial field, due to the lack of relevant training data, it is often necessary to manually write training corpus and spend a lot of time training natural language processing models such as Bert. However, compared with the general language model, Bert has a low recognition accuracy and has limitations in semantic understanding.

[0060] In this regard, after obtaining the target financial industry knowledge text, since the knowledge text has a semantic association with the text to be recognized input by the user, understanding the target financial industry knowledge text will help understand the intention of the user inputting the text to be recognized. In this embodiment, the target financial industry knowledge text and the text to be recognized can be input into a pre-trained general language model. In some optional embodiments, the pre-trained general language model can be an existing large-scale language model (LLM). The general language model then analyzes the target financial industry knowledge text to understand the content of the vertical field (i.e., the financial field) associated with the text to be recognized, so that the general language model has not only the knowledge of the general language model, but also the knowledge of the vertical field. After that, the context semantics of the text to be recognized can be determined in combination with the target financial industry knowledge text, and the user intention corresponding to the text to be recognized can be determined based on the context semantics.

[0061] In this embodiment, the text to be recognized that is input by the user in the financial service application and to be recognized for intent recognition can be obtained, and then the text features corresponding to the text to be recognized are compared with the text features corresponding to multiple financial industry knowledge texts in the pre-constructed financial industry knowledge base, and the target financial industry knowledge text associated with the text to be recognized is determined according to the text feature similarity obtained by the comparison, and then the target financial industry knowledge text and the text to be recognized can be input into a pre-trained general language model, and the general language model determines the contextual semantics of the text to be recognized according to the target financial industry knowledge text, and determines the user intent corresponding to the text to be recognized according to the contextual semantics. In this embodiment, by inputting the target financial industry knowledge text that has a language association with the text to be recognized into the general language model, it is possible to avoid secondary training of the general language model, so that the language model can be adapted to the financial field and understand the target financial industry knowledge, thereby helping to quickly understand the user intent of the user who inputs the text to be recognized in the financial service application, and improve the efficiency of intent recognition.

[0062] In one embodiment, S102 compares the text features corresponding to the text to be recognized with the text features corresponding to multiple financial industry knowledge texts in a pre-built financial industry knowledge base, which may include the following steps:

[0063] The text features corresponding to the text to be recognized are compared with the text features corresponding to each financial industry knowledge text in the pre-built financial industry knowledge base to obtain the text feature similarity between the text to be recognized and each financial industry knowledge text.

[0064] In practical applications, the financial industry knowledge texts stored in the financial industry knowledge base can be continuously adjusted and updated, for example, the financial industry knowledge in the financial industry knowledge base can be deleted or added. In this embodiment, the text features corresponding to the text to be identified can be compared with the text features corresponding to each financial industry knowledge text in the pre-built financial industry knowledge base to obtain the text feature similarity between the text to be identified and each financial industry knowledge text. By comparing the text features of the text to be identified with the text features corresponding to each pre-stored financial industry knowledge text, it can be ensured that the relevant target financial industry knowledge text is accurately found in the continuously updated and adjusted financial industry knowledge base.

[0065] In one embodiment, Figure 2 As shown, S102 compares the text features corresponding to the text to be recognized with the text features corresponding to multiple financial industry knowledge texts in the pre-built financial industry knowledge base, which may include the following steps:

[0066] S201, determining knowledge labels corresponding to various categories of financial industry knowledge in a pre-built financial industry knowledge base.

[0067] In practical applications, the financial industry knowledge texts stored in the pre-built financial industry knowledge base can be divided into multiple categories to obtain multiple categories of financial industry knowledge. For each category of financial industry knowledge, a corresponding knowledge label can be set, and each category of financial industry knowledge can have at least one knowledge label.

[0068] Furthermore, after obtaining the text to be recognized, the knowledge tags corresponding to each category of financial industry knowledge can be determined.

[0069] S202, matching the text to be recognized with each knowledge tag, and determining the target category financial industry knowledge corresponding to the matched knowledge tag.

[0070] In this step, the text to be recognized may be matched with each knowledge tag to determine the matching knowledge tag, and the financial industry knowledge corresponding to the matching knowledge tag may be used as the target category financial industry knowledge.

[0071] In some optional embodiments, the text to be recognized may be segmented, and matching keywords may be determined based on multiple text segmentations obtained from the segmentation process, and then the matching keywords may be matched with various knowledge tags.

[0072] S203, comparing text features corresponding to the text to be recognized with text features corresponding to the financial industry knowledge text of the target category of financial industry knowledge.

[0073] After the target category of financial industry knowledge is determined, each financial industry knowledge text of the target category of financial industry knowledge may be obtained, and the text features corresponding to the text to be identified may be compared with the text features of each obtained financial industry knowledge text.

[0074] In this embodiment, by matching the text to be identified with the knowledge label, the target category of financial industry knowledge can be determined, so that the text features of the text to be identified are only compared with the text features corresponding to the financial industry knowledge text of the target category of financial industry knowledge, accurately narrowing the comparison scope and improving the comparison efficiency.

[0075] In one embodiment, the financial industry knowledge texts stored in the financial industry knowledge base can also be counted to obtain the total number of financial industry knowledge texts. Since the financial industry knowledge texts can be in a state of continuous adjustment and updating, after obtaining the text to be recognized, it can be determined whether the total number of current financial industry knowledge texts is greater than a threshold. If so, steps S201-S203 are executed; if not, the step of comparing the text features corresponding to the text to be recognized with the text features corresponding to each financial industry knowledge text in the pre-built financial industry knowledge base is executed.

[0076] In one embodiment, Figure 3 As shown in the figure, the financial industry knowledge base is constructed through the following steps:

[0077] S301, clustering a plurality of pre-acquired financial industry knowledge texts to obtain a plurality of categories of financial industry knowledge and financial industry knowledge texts corresponding to each category of financial industry knowledge.

[0078] In a specific implementation, multiple pre-acquired financial industry knowledge texts can be clustered, for example, using the kmeans algorithm to obtain multiple clusters, each cluster including at least one financial industry knowledge text, and each cluster corresponding to a category of financial industry knowledge, thereby obtaining multiple categories of financial industry knowledge and financial industry knowledge texts corresponding to each category of financial industry knowledge.

[0079] S302, for each category of financial industry knowledge, obtain a text vector corresponding to each financial industry knowledge text of the category of financial industry knowledge, and determine a text feature corresponding to the financial industry knowledge text of the category of financial industry knowledge based on multiple text vectors.

[0080] For each category of financial industry knowledge, the text vectors corresponding to each financial industry knowledge text of the category of financial industry knowledge can be obtained, and the text features corresponding to the financial industry knowledge text of the category of financial industry knowledge can be determined based on multiple text vectors. For example, multiple text vectors can be used as text features corresponding to the financial industry knowledge text of the category of financial industry knowledge, or multiple text vectors can be fused and the fused text vector can be used as the text features corresponding to the financial industry knowledge text of the category of financial industry knowledge.

[0081] In this embodiment, by acquiring and classifying multiple financial industry knowledge texts, it is possible to quickly perform targeted comparisons when performing text feature comparisons, thereby improving the efficiency of intent recognition.

[0082] In one embodiment, determining the target financial industry knowledge text associated with the to-be-recognized text according to the text feature similarity obtained by comparison in step S102 may include the following steps:

[0083] Determine the difficulty of intent recognition of the text to be recognized, and determine the number K of text screening corresponding to the difficulty of intent recognition; based on the text feature similarity obtained by comparison, obtain the K financial industry knowledge texts with the highest text feature similarity from multiple financial industry knowledge texts to obtain the financial industry knowledge text associated with the text to be recognized.

[0084] Among them, K is a positive integer, and K is positively correlated with the difficulty of intent recognition.

[0085] In a specific implementation, the contents of different texts to be recognized may differ. For example, the length of the text to be recognized and the number of valid keywords contained in the text to be recognized may differ. Accordingly, the difficulty of understanding the intent may also differ for different texts to be recognized.

[0086] In this embodiment, the difficulty of intent recognition of the text to be recognized can be determined, and the number of text screening K corresponding to the difficulty of intent recognition can be determined. In some optional embodiments, it can be determined according to the text length of the text to be recognized or the number of valid keywords contained therein, for example, the shorter the text length, the higher the difficulty of intent recognition, and the fewer the number of valid keywords, the higher the difficulty of intent recognition, wherein the valid keywords can refer to the segmentation that assists the general language model in determining the user's intent.

[0087] Furthermore, based on the text feature similarity obtained by comparison, K financial industry knowledge texts with the highest text feature similarity can be obtained from multiple financial industry knowledge texts to obtain the financial industry knowledge text associated with the text to be identified.

[0088] In this embodiment, by obtaining the K financial industry knowledge texts with the highest text feature similarity as the financial industry knowledge texts associated with the text to be recognized, it is possible to add a corresponding number of financial industry knowledge texts to the general language model according to changes in the difficulty of intent recognition of the text to be recognized, so that the general language model can understand the corresponding information and improve the accuracy of intent recognition.

[0089] In one embodiment, S101 obtains text to be recognized that is input by a user in a financial service application and to be recognized by intent, and may include the following steps:

[0090] The text information entered by the user in the search box of the financial services application is used as the text to be recognized for intent recognition; and / or the text information entered by the user in the dialog box of the financial services application is used as the text to be recognized for intent recognition.

[0091] In this embodiment, by using the text information entered by the user in the search box as the text to be recognized, matching search results can be returned to the user according to the user's intention; or, by using the text information entered in the dialog box as the text to be recognized, the chatbot in the financial services application can provide correct feedback information when the user has a conversation with the chatbot.

[0092] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0093] Based on the same inventive concept, the embodiment of the present application also provides a device for identifying the intention of a financial service application for implementing the method for identifying the intention of a financial service application involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the embodiments of the device for identifying the intention of one or more financial service applications provided below can refer to the limitations of the method for identifying the intention of a financial service application above, and will not be repeated here.

[0094] In one embodiment, Figure 4 As shown, a device for identifying intentions of financial service applications is provided, comprising:

[0095] The text acquisition module 401 is used to acquire the text to be recognized that is input by the user in the financial service application and is to be recognized by the intent;

[0096] A feature comparison module 402 is used to compare the text features corresponding to the text to be recognized with the text features corresponding to multiple financial industry knowledge texts in a pre-built financial industry knowledge base, and determine the target financial industry knowledge text associated with the text to be recognized based on the text feature similarity obtained by the comparison;

[0097] The intention recognition module 403 is used to input the target financial industry knowledge text and the text to be recognized into a pre-trained general language model, and the general language model determines the contextual semantics of the text to be recognized based on the target financial industry knowledge text, and determines the user intention corresponding to the text to be recognized based on the contextual semantics.

[0098] In one embodiment, the feature comparison module 402 is used to:

[0099] The text features corresponding to the text to be recognized are compared with the text features corresponding to each financial industry knowledge text in the pre-built financial industry knowledge base to obtain the text feature similarity between the text to be recognized and each financial industry knowledge text.

[0100] In one embodiment, the feature comparison module 402 is used to:

[0101] Determine the knowledge labels corresponding to each category of financial industry knowledge in the pre-built financial industry knowledge base;

[0102] Matching the text to be recognized with each knowledge tag, and determining the target category financial industry knowledge corresponding to the matched knowledge tag;

[0103] The text features corresponding to the text to be recognized are compared with the text features corresponding to the financial industry knowledge text of the target category financial industry knowledge.

[0104] In one embodiment, the apparatus further comprises:

[0105] A clustering module is used to cluster multiple pre-acquired financial industry knowledge texts to obtain multiple categories of financial industry knowledge and financial industry knowledge texts corresponding to each category of financial industry knowledge;

[0106] The knowledge base construction module is used to obtain the text vector corresponding to each financial industry knowledge text of each category of financial industry knowledge, and determine the text features corresponding to the financial industry knowledge text of the category of financial industry knowledge based on multiple text vectors.

[0107] In one embodiment, the feature comparison module 402 is used to:

[0108] Determine the difficulty of intention recognition of the text to be recognized, and determine the number K of text screening corresponding to the difficulty of intention recognition; K is a positive integer, and K is positively correlated with the difficulty of intention recognition;

[0109] According to the text feature similarity obtained by comparison, K financial industry knowledge texts with the highest text feature similarity are obtained from the multiple financial industry knowledge texts to obtain the financial industry knowledge text associated with the text to be identified.

[0110] In one embodiment, the text acquisition module 401 is used to:

[0111] The text information entered by the user in the search box of the financial service application is used as the text to be recognized for intent recognition;

[0112] and / or,

[0113] The text information input by the user in the dialog box of the financial service application is used as the text to be recognized for intent recognition.

[0114] Each module in the above-mentioned intention recognition device for financial service applications can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.

[0115] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data for implementing an intention recognition method for financial service applications. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for identifying the intention of a financial service application is implemented.

[0116] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 6 As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a method for identifying the intention of a financial service application is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0117] Those skilled in the art will understand that Figure 5 and Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0118] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0119] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0120] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0121] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0122] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0123] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0124] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for identifying intent in a financial service application, characterized in that: The method comprises: Obtaining text to be recognized that is input by a user in a financial service application and to be recognized by intent; Comparing the text features corresponding to the text to be recognized with the text features corresponding to multiple financial industry knowledge texts in a pre-built financial industry knowledge base, and determining the target financial industry knowledge text associated with the text to be recognized based on the text feature similarity obtained by the comparison; The target financial industry knowledge text and the text to be recognized are input into a pre-trained general language model, and the general language model determines the contextual semantics of the text to be recognized based on the target financial industry knowledge text, and determines the user intention corresponding to the text to be recognized based on the contextual semantics.

2. The method according to claim 1, characterized in that The comparing the text features corresponding to the text to be recognized with the text features corresponding to a plurality of financial industry knowledge texts in a pre-built financial industry knowledge base includes: The text features corresponding to the text to be recognized are compared with the text features corresponding to each financial industry knowledge text in the pre-built financial industry knowledge base to obtain the text feature similarity between the text to be recognized and each financial industry knowledge text.

3. The method according to claim 1, characterized in that: The comparing the text features corresponding to the text to be recognized with the text features corresponding to a plurality of financial industry knowledge texts in a pre-built financial industry knowledge base includes: Determine the knowledge labels corresponding to each category of financial industry knowledge in the pre-built financial industry knowledge base; Matching the text to be recognized with each knowledge tag, and determining the target category financial industry knowledge corresponding to the matched knowledge tag; The text features corresponding to the text to be recognized are compared with the text features corresponding to the financial industry knowledge text of the target category financial industry knowledge.

4. The method according to claim 3, characterized in that The financial industry knowledge base is constructed through the following steps: Clustering multiple pre-acquired financial industry knowledge texts to obtain multiple categories of financial industry knowledge and financial industry knowledge texts corresponding to each category of financial industry knowledge; For each category of financial industry knowledge, a text vector corresponding to each financial industry knowledge text of the category of financial industry knowledge is obtained, and based on multiple text vectors, a text feature corresponding to the financial industry knowledge text of the category of financial industry knowledge is determined.

5. The method according to claim 1, characterized in that Determining the target financial industry knowledge text associated with the to-be-recognized text according to the text feature similarity obtained by comparison includes: Determine the difficulty of intention recognition of the text to be recognized, and determine the number K of text screening corresponding to the difficulty of intention recognition; K is a positive integer, and K is positively correlated with the difficulty of intention recognition; According to the text feature similarity obtained by comparison, K financial industry knowledge texts with the highest text feature similarity are obtained from the multiple financial industry knowledge texts to obtain the financial industry knowledge text associated with the text to be identified.

6. The method according to any one of claims 1 to 5, characterized in that: The step of obtaining the text to be recognized that is input by the user in the financial service application and to be recognized by the intent includes: The text information entered by the user in the search box of the financial service application is used as the text to be recognized for intent recognition; and / or, The text information input by the user in the dialog box of the financial service application is used as the text to be recognized for intent recognition.

7. An intention recognition device for financial service applications, characterized in that: The device comprises: A text acquisition module, used to acquire text to be recognized that is input by a user in a financial service application and is to be recognized by intent; A feature comparison module is used to compare the text features corresponding to the text to be identified with the text features corresponding to multiple financial industry knowledge texts in a pre-built financial industry knowledge base, and determine the target financial industry knowledge text associated with the text to be identified based on the text feature similarity obtained by the comparison; The intention recognition module is used to input the target financial industry knowledge text and the text to be recognized into a pre-trained general language model, and the general language model determines the contextual semantics of the text to be recognized based on the target financial industry knowledge text, and determines the user intention corresponding to the text to be recognized based on the contextual semantics.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images