Fraud intention recognition method and device, electronic equipment and storage medium

Abstract

The invention provides a fraud intention recognition method and device, electronic equipment and a storage medium. The method comprises the steps of firstly obtaining a to-be-detected text; then respectively inputting the to-be-detected text into a fraud mode discrimination model, an emotion analysis model and a keyword extraction model obtained by training to obtain a fraud mode, an emotion category and keywords corresponding to the to-be-detected text; and finally, inputting the to-be-detected text, the fraud mode, the emotion category and the keyword into a fraud intention recognition model obtained by training to obtain a fraud intention recognition result of the to-be-detected text. Therefore, multi-dimensional analysis of text content can be realized through the fraud mode discrimination model, the sentiment analysis model and the keyword extraction model, the fraud mode, the sentiment category and the keyword are obtained, the fraud intention is comprehensively judged in combination with the to-be-detected text, and the recognition accuracy and efficiency of novel fraud behaviors are effectively improved.

Description

Technical Field

[0001] The present application relates to the technical field of fraudulent intent identification, and in particular to a fraudulent intent identification method, device, electronic device and storage medium. Background Art

[0002] With the rapid development of the Internet and communication technology, online fraud methods are becoming increasingly diverse and complex. Fraudsters use modern technology to disguise themselves as legitimate identities and conduct fraud through text messages, social media, emails, and other channels. These fraud methods have not only increased in number, but also become increasingly complex in technology, making it difficult for traditional fraud prevention methods to effectively deal with them. For example, traditional rule-based anti-fraud systems often rely on pre-set features and keywords for detection, but their effectiveness gradually decreases in the face of new and mutated fraud methods. In addition, fraud detection methods based on single-dimensional analysis, such as relying only on semantic analysis or keyword extraction, also have detection blind spots and are difficult to deal with the diversity and complexity of fraud methods. Summary of the invention

[0003] In view of this, the purpose of this application is to propose a method, device, electronic device and storage medium for identifying fraudulent intent to solve or partially solve the above-mentioned technical problems.

[0004] Based on the above purpose, this application provides a method for identifying fraudulent intentions, including:

[0005] Get the text to be detected;

[0006] The text to be detected is respectively input into the fraud pattern discrimination model, sentiment analysis model and keyword extraction model obtained through training to obtain the fraud pattern, sentiment category and keyword corresponding to the text to be detected;

[0007] The text to be detected, the fraud pattern, the emotion category and the keyword are input into the fraud intention recognition model obtained through training to obtain the fraud intention recognition result of the text to be detected.

[0008] Based on the same concept, the present application also proposes a fraud intention identification device, comprising:

[0009] Get module to get the text to be detected;

[0010] The prompt word module inputs the text to be detected into the fraud pattern discrimination model, sentiment analysis model and keyword extraction model obtained through training, respectively, to obtain the fraud pattern, sentiment category and keyword corresponding to the text to be detected;

[0011] The recognition module inputs the text to be detected, the fraud pattern, the emotion category and the keyword into the fraud intention recognition model obtained through training to obtain the fraud intention recognition result of the text to be detected.

[0012] Based on the same concept, the present application also proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the fraudulent intent identification method as described above when executing the program.

[0013] Based on the same concept, the present application also proposes a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable a computer to execute the fraud intention identification method as described above.

[0014] Based on the same concept, the present application also proposes a computer program product, which includes a computer program. The computer program is executed by one or more processors so that the processors execute the fraud intention identification method as described above.

[0015] From the above, it can be seen that the fraud intention identification method, device, electronic device and storage medium provided by the present application first obtain the text to be detected; then the text to be detected is respectively input into the trained fraud pattern discrimination model, sentiment analysis model and keyword extraction model to obtain the fraud pattern, sentiment category and keyword corresponding to the text to be detected; finally, the text to be detected, the fraud pattern, the sentiment category and the keyword are input into the trained fraud intention identification model to obtain the fraud intention identification result of the text to be detected, so that the fraud pattern discrimination model, sentiment analysis model and keyword extraction model can realize multi-dimensional analysis of text content, obtain fraud pattern, sentiment category and keywords, and make a comprehensive judgment on fraud intention in combination with the text to be detected, thereby effectively improving the accuracy and efficiency of identifying new fraud behaviors. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the present application or related technologies, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 A schematic diagram of an application scenario of an embodiment of the present application;

[0018] Figure 2 A flowchart of a fraud intention identification method according to an embodiment of the present application;

[0019] Figure 3 A schematic diagram of the structure of a fraud intention identification device according to an embodiment of the present application;

[0020] Figure 4 A schematic diagram of a specific electronic device hardware structure of an embodiment of the present application. DETAILED DESCRIPTION

[0021] It is understandable that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and relevant provisions.

[0022] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present application should be the usual meanings understood by people with general skills in the field to which the present application belongs. The "first", "second" and similar words used in the present application do not represent any order, quantity or importance, but are only used to distinguish different components. "Including" or "comprising" and similar words mean that the elements or objects appearing in front of the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connecting" or "connected" and similar words are not limited to physical or mechanical connections, but can include electrical connections, which can be direct connections, indirect connections, wired connections, and wireless connections. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly. In addition, in the description of the present application, unless otherwise specified, the term "multiple" refers to two or more. The term "and / or" describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects are in an "or" relationship.

[0023] The principles and spirit of the present application will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and implement the present application, and are not intended to limit the scope of the present application in any way. On the contrary, these embodiments are provided to make the present application more thorough and complete, and to fully convey the scope of the present application to those skilled in the art.

[0024] It is understandable that before using the technical solutions of each embodiment of the present application, the type, scope of use, usage scenarios, etc. of the personal information involved will be informed to the user in an appropriate manner, and the user's authorization will be obtained.

[0025] For example, in response to receiving an active request from a user, a prompt message is sent to the user to clearly remind the user that the operation requested to be performed will require obtaining and using the user's personal information. Thus, the user can independently choose whether to provide personal information to the electronic device, application, server, storage medium or other software or hardware that performs the operation of the technical solution of the present application according to the prompt message.

[0026] As an optional but non-limiting implementation, in response to receiving the user's active request, the prompt information may be sent to the user in the form of a pop-up window, in which the prompt information may be presented in text form. In addition, the pop-up window may also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0027] It is understandable that the above notification and the process of obtaining user authorization are merely illustrative and do not constitute a limitation on the implementation method of this application. Other methods that meet relevant laws and regulations may also be applied to the implementation method of this application.

[0028] It should be understood herein that any number of elements in the drawings is for illustration rather than limitation, and any naming is only for distinction rather than having any limiting meaning.

[0029] As described in the background technology section, with the rapid development of the Internet and communication technology, online fraud methods are becoming increasingly diverse and complex. Fraudsters use modern technology to disguise themselves as legitimate identities and conduct fraud through text messages, social media, emails, and other channels. These fraud methods have not only increased in number, but also become increasingly complex in technology, making it difficult for traditional fraud prevention methods to effectively respond. For example, traditional rule-based anti-fraud systems often rely on pre-set features and keywords for detection, but their effectiveness gradually decreases in the face of new and mutated fraud methods. In addition, fraud detection methods based on single-dimensional analysis, such as relying only on semantic analysis or keyword extraction, also have detection blind spots, which are difficult to cope with the diversity and complexity of fraud methods.

[0030] In recent years, with the development of artificial intelligence, especially AI model technology, text analysis technology based on deep learning has gradually been applied to the field of network security. Artificial intelligence models have strong natural language understanding capabilities, can capture complex language patterns, and show high accuracy in a variety of tasks. However, although artificial intelligence models have achieved remarkable results in natural language processing tasks, the direct application of artificial intelligence models for fraud intent identification still faces challenges. This is because the identification of fraud intent not only depends on the text content itself, but also involves the comprehensive analysis of multi-dimensional information such as fraud patterns, emotional tendencies, keywords, etc. Traditional models can often only perform well in one dimension and cannot effectively integrate information from multiple dimensions, resulting in limitations in the recognition effect.

[0031] Therefore, how to take advantage of multi-dimensional prompt words and AI models to accurately identify potential fraudulent intentions by comprehensively analyzing the fraud patterns, emotional tendencies, keywords and other information of the text to be detected has become an urgent problem to be solved in the current anti-fraud technology field. In addition, as fraud methods are constantly updated, how the system can maintain its sensitivity and accurate recognition capabilities to new fraud methods is also one of the technical difficulties in this field.

[0032] In summary, in order to solve the above problems, the present application embodiment proposes a fraud intention identification method, including:

[0033] First, the text to be detected is obtained; then the text to be detected is respectively input into the fraud pattern discrimination model, sentiment analysis model and keyword extraction model obtained through training, so as to obtain the fraud pattern, sentiment category and keyword corresponding to the text to be detected; finally, the text to be detected, the fraud pattern, the sentiment category and the keyword are input into the fraud intention recognition model obtained through training, so as to obtain the fraud intention recognition result of the text to be detected, thereby realizing multi-dimensional analysis of text content through the fraud pattern discrimination model, sentiment analysis model and keyword extraction model, obtaining the fraud pattern, sentiment category and keywords, and making a comprehensive judgment on the fraud intention in combination with the text to be detected, so as to effectively improve the recognition accuracy and efficiency of new fraud behaviors.

[0034] In some specific application scenarios, the fraudulent intention identification method of the present application can be applied to various platforms or systems involving fraudulent intention identification. As an example, the application scenario includes at least one server and at least one terminal. Figure 1, which is a schematic diagram of an application scenario of a fraud intention identification method provided by an exemplary embodiment of the present disclosure. The application scenario includes a terminal device 101, a server 102, and a data storage system 103. Among them, the terminal device 101, the server 102 and the data storage system 103 can be connected through a wired or wireless communication network. The terminal device includes but is not limited to a desktop computer, a mobile phone, a mobile computer, a tablet computer, a media player, a smart wearable device, a personal digital assistant (PDA) or other electronic devices capable of realizing the above functions. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server and the terminal can communicate through the network to realize data transmission. Among them, the network can be a wired network or a wireless network, and this application does not make specific restrictions on this.

[0035] The server may be a server that provides various services. Specifically, the server may be used to provide background services for applications running on the terminal. Optionally, in some implementations, the fraudulent intent identification method provided in the embodiment of the present application may be executed by the server. Optionally, in some implementations, the fraudulent intent identification method provided in the embodiment of the present application may be directly executed by the terminal device.

[0036] Optionally, the server can be hardware or software. When the server is hardware, it can be implemented as a distributed server cluster consisting of multiple servers, or it can be implemented as a single server. When the server is software, it can be implemented as multiple software or software modules (for example, software or software modules for providing distributed services), or it can be implemented as a single software or software module. This embodiment of the application does not specifically limit this.

[0037] Optionally, the wireless network or wired network uses standard communication technology and / or protocol. The network is usually the Internet, but it can also be any network, including but not limited to a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a mobile, wired or wireless network, a dedicated network or any combination of a virtual private network). In some embodiments, the data exchanged through the network is represented by technologies and / or formats including hypertext mark-up language (HTML), extensible markup language (XML), etc. In addition, conventional encryption technologies such as secure socket layer (SSL), transport layer security (TLS), virtual private network (VPN), and Internet protocol security (IPsec) can be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technologies can also be used to replace or supplement the above data communication technologies.

[0038] The data storage system stores a large amount of data. Optionally, the data storage system can partition the data according to different uses and data types. The source of the data includes but is not limited to an existing database, data crawled from the Internet, or data uploaded by the user when using the client.

[0039] The following describes the fraudulent intention identification method according to the exemplary embodiment of the present application in conjunction with a specific application scenario. It should be noted that the above application scenario is only shown to facilitate understanding of the spirit and principle of the present application, and the implementation of the present application is not limited in this regard. On the contrary, the implementation of the present application can be applied to any applicable scenario.

[0040] refer to Figure 2 , is a flow chart of a method for identifying fraudulent intent according to an embodiment of the present application, the method comprising the following steps:

[0041] S101, obtaining the text to be detected.

[0042] In specific implementation, the acquired text to be detected can be obtained by user input, or transmitted through other devices or systems, and there is no limitation on this. The text to be detected is the text that currently needs to be identified for fraudulent intent.

[0043] S102, inputting the text to be detected into the fraud pattern discrimination model, sentiment analysis model and keyword extraction model obtained through training respectively, to obtain the fraud pattern, sentiment category and keyword corresponding to the text to be detected.

[0044] In specific implementation, after obtaining the text to be detected, the trained fraud pattern discrimination model, sentiment analysis model and keyword extraction model can be called in sequence to parse the text from different dimensions to obtain the fraud pattern, sentiment category and keywords of the text.

[0045] It should be noted that in this process, each model works independently and completes the analysis task of the text. For example, in some embodiments, first, the text to be detected is input into the fraud pattern discrimination model, and the output result is one or more fraud pattern labels, which represent the potential fraud methods in the text identified by the system. Next, the text to be detected is input into the sentiment analysis model, and the output result is the sentiment category label, which is used to represent the emotional response that the fraud text may trigger. Finally, the text to be detected is input into the keyword extraction model, and the output result is a group of keywords, which are closely related to the fraud methods and help to further determine the specific intentions and risk levels of the fraud. Optionally, in order to improve efficiency, fraud pattern discrimination, sentiment analysis and keyword extraction can be performed simultaneously in a parallel processing manner to reduce the overall detection time.

[0046] In some embodiments, the fraud mode includes: one or more of an identity impersonation fraud mode, a false investment fraud mode, and a lottery fraud mode; the emotion category includes: one or more of fear, tension, anxiety, and excitement; the keywords include: one or more of a bank account, remittance, bonus, and password.

[0047] It should be noted that in this embodiment, three models are constructed for the fraud pattern discrimination task, sentiment classification task, and keyword extraction task, including a fraud pattern discrimination model, a sentiment analysis model, and a keyword extraction model, which will be used to parse the text to be tested from different dimensions. Optionally, the above three models can be AI large models. The specific process of training the model can refer to the relevant technology, and the training process of the neural network model is not limited.

[0048] In some embodiments, before inputting the to-be-detected text into the trained fraud pattern discrimination model, sentiment analysis model, and keyword extraction model, respectively, the method further includes:

[0049] Get historical fraud text data;

[0050] Classify the historical fraud text data to obtain a fraud pattern discrimination data set, a sentiment analysis data set, and a keyword extraction data set;

[0051] Training a first benchmark model based on the fraud pattern discrimination data set to obtain the fraud pattern discrimination model;

[0052] Training a second benchmark model based on the sentiment analysis data set to obtain the sentiment analysis model;

[0053] A third benchmark model is trained based on the keyword extraction data set to obtain the keyword extraction model.

[0054] In the specific implementation, in order to accurately train the model, it is necessary to first obtain historical fraud text data, classify the historical fraud text data, and establish multiple different types of data sets related to fraud. Optionally, the data set includes at least a fraud pattern discrimination data set, a sentiment analysis data set, and a keyword extraction data set. Optionally, when obtaining historical fraud text data, a large amount of historical fraud text data can be collected from multiple channels. These channels may include fraud text messages, telephone recording texts, etc. After obtaining the historical fraud text data, it can be classified by data annotation, or the data classification can be completed by an AI classification model without limitation. After obtaining the fraud pattern discrimination data set, sentiment analysis data set, and keyword extraction data set, the above three data sets can be used to train the fraud pattern discrimination model, sentiment analysis model, and keyword extraction model respectively. It should be noted that the above-mentioned first benchmark model, second benchmark model, and third benchmark model can be selected as needed, without limitation.

[0055] Optionally, in some embodiments, the labeling process involves adding a label to each text data according to a preset standard, specifically including:

[0056] Fraud pattern identification and annotation: Classify and annotate each text according to the fraud method, such as impersonation fraud, false investment fraud, false lottery fraud, etc.

[0057] Sentiment analysis annotation: Annotate the emotional categories that fraudulent texts may trigger, such as fear, anxiety, tension, excitement, etc.

[0058] Keyword extraction and annotation: Extract fraud-related keywords or phrases from the text, such as "bank account", "transfer", "bonus", "password", etc.

[0059] Based on these labeled data, we established fraud pattern discrimination datasets, sentiment analysis datasets, and keyword extraction datasets. Each dataset corresponds to a different analysis dimension, providing data support for subsequent model training and prompt word generation.

[0060] In some embodiments, when constructing a fraud pattern discrimination model, a deep learning framework for natural language processing can be used, such as a model based on a Transformer architecture (e.g., BERT, GPT, etc.). During the training process, the model can automatically identify potential fraudulent methods in the text by learning the language pattern of the fraudulent text. The model continuously optimizes parameters through multiple rounds of iterative training to ensure accurate classification of different fraud patterns. The construction of the sentiment analysis model is based on a sentiment analysis data set. The model can adopt a deep learning method commonly used in sentiment classification tasks, such as LSTM (long short-term memory network) or a pre-trained Transformer model. By learning the emotional features of the fraudulent text, the sentiment analysis model can effectively identify the potential emotional tendencies in the text and help the system determine the psychological impact that the fraudulent text may have on the victim. Optionally, in model training, methods such as cross-validation can be used to ensure the model's ability to distinguish emotional categories. The constructed keyword extraction model can be based on a common framework for sequence labeling tasks, such as BiLSTM-CRF (bidirectional long short-term memory network combined with conditional random field) or BERT-CRF model. By learning the association between the contextual information of the text and the keywords, the model can accurately identify keywords with fraud risks in the text.

[0061] Optionally, in some embodiments, when training the above three models, the models can use labeled data for supervised learning. In order to improve the generalization ability and accuracy of the model, strategies such as data enhancement, transfer learning, and hyperparameter optimization can be used during the training process.

[0062] In some embodiments, before classifying the historical fraud text data, the method further includes:

[0063] The historical fraud text data is cleaned; wherein the data cleaning step includes: removing duplicate data, correcting format problems in the text, filtering out one or more of preset irrelevant information.

[0064] In specific implementation, before classifying the historical fraud text data, the collected raw text data can be cleaned to remove noise and improve data quality. The cleaning process includes but is not limited to the following operations: removing duplicate data, correcting format problems in the text (such as character encoding errors or incomplete sentences), filtering out irrelevant information (such as non-fraudulent content or low-quality junk data), etc.

[0065] S103, inputting the text to be detected, the fraud pattern, the emotion category and the keyword into the fraud intention recognition model obtained through training to obtain a fraud intention recognition result of the text to be detected.

[0066] In specific implementation, after obtaining the fraud pattern, sentiment category and keywords corresponding to the text to be detected, the multi-dimensional analysis results of the text to be detected are obtained. These analysis results independently reflect the different feature dimensions of the text, and these analysis results are input into the trained fraud intent recognition model together with the original text to be detected. The fraud intent recognition model combines the multi-dimensional information of the original text and the prompt words, analyzes the overall context and content of the text, determines whether there is fraud intent, and obtains the fraud intent recognition result of the text to be detected, thereby further ensuring the accuracy and efficiency of fraud intent recognition. It should be noted that the process of training the fraud intent recognition model can refer to the relevant technology, and there is no limitation on the training of the neural network model.

[0067] In some embodiments, in order to further improve the accuracy of the fraud intent recognition model, after obtaining the fraud pattern, the emotion category and the keyword, these independent results can be integrated to generate a multidimensional prompt word, and then the generated multidimensional prompt word can be input into the fraud intent recognition model together with the text to be detected.

[0068] In some embodiments, the fraud patterns, sentiment categories, and keywords generated by a text to be detected specifically include:

[0069] Fraud pattern: Identity fraud;

[0070] Emotional categories: nervous;

[0071] Keywords: remittance, password, bank account.

[0072] In some embodiments, the method further comprises:

[0073] Obtain target fraud text data; the target fraud text data is fraud text data collected after a preset time interval of completing training of the fraud pattern discrimination model, the sentiment analysis model, and the keyword extraction model;

[0074] The parameters of the fraud pattern discrimination model, the sentiment analysis model and the keyword extraction model are updated based on the target fraud text data.

[0075] During specific implementation, in order to ensure that the model can continue to effectively respond to ever-changing fraud methods and maintain sensitivity and high accuracy to new types of fraud, in this embodiment, the parameters of the fraud pattern discrimination model, the sentiment analysis model and the keyword extraction model are updated by regularly collecting target fraud text data.

[0076] In some embodiments, a model update mechanism based on user feedback and the latest fraud cases is also designed. The mechanism mainly includes the collection of user feedback, the accumulation of new fraud cases and the regular update of each model.

[0077] Users can provide feedback in several ways:

[0078] Feedback on false positives or missed positives: If users find that the fraud recognition model’s recognition results for certain fraudulent texts are inaccurate (such as misjudging them as normal or missing fraudulent intent) during use, they can provide feedback on these false positives or missed positives to the system.

[0079] User annotation feedback: Users can manually correct or confirm the recognition results of the fraud recognition model and use the annotation information as reference data for model optimization.

[0080] In some embodiments, the latest fraud text (target fraud text data) can be obtained through the following ways:

[0081] Real-time monitoring of data sources to capture new fraud-related text data from the Internet, social media, email and other channels in real time;

[0082] Update external data sets, combining the latest fraud sample sets released by authoritative institutions or security platforms to expand the system's fraud text library;

[0083] New cases submitted by users. Users can upload or mark the latest fraud information, and the system will include these user-submitted cases into the fraud text library.

[0084] In some embodiments, after accumulating enough latest fraud texts and user feedback, the new data can be processed through the above-mentioned data cleaning steps to obtain an expanded fraud pattern discrimination data set, sentiment analysis data set and keyword extraction data set. Based on the expanded data set, the fraud pattern discrimination model, sentiment analysis model and keyword extraction model are retrained. Optionally, in some embodiments, transfer learning technology can be introduced to reduce training time and improve the generalization ability of the new model. During the update process, the model's hyperparameter adjustment and architecture optimization will be tuned according to the latest data performance to improve the performance of the model. After the update of each sub-model (fraud pattern, sentiment analysis, keyword extraction) is completed, the fraud intent recognition model can also be updated and trained according to the latest multi-dimensional prompt words.

[0085] The fraud intention identification method provided by the present application first obtains the text to be detected; then the text to be detected is respectively input into the fraud pattern discrimination model, sentiment analysis model and keyword extraction model obtained by training, and the fraud pattern, sentiment category and keyword corresponding to the text to be detected are obtained; finally, the text to be detected, the fraud pattern, the sentiment category and the keyword are input into the fraud intention identification model obtained by training, and the fraud intention identification result of the text to be detected is obtained, so that the fraud pattern discrimination model, sentiment analysis model and keyword extraction model can realize multi-dimensional analysis of text content, obtain fraud patterns, sentiment categories and keywords, and make a comprehensive judgment of fraud intentions in combination with the text to be detected, effectively improving the recognition accuracy and efficiency of new fraud behaviors. In addition, the present application significantly improves the accuracy of fraud intention identification and enhances the security protection capabilities for users.

[0086] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides a fraud intention identification device, which is applied to the first server.

[0087] refer to Figure 3 , the fraud intention identification device comprises:

[0088] Acquisition module 201, acquiring the text to be detected;

[0089] The prompt word module 202 inputs the text to be detected into the fraud pattern discrimination model, sentiment analysis model and keyword extraction model obtained through training, respectively, to obtain the fraud pattern, sentiment category and keyword corresponding to the text to be detected;

[0090] The recognition module 203 inputs the text to be detected, the fraud pattern, the emotion category and the keyword into the fraud intention recognition model obtained through training to obtain the fraud intention recognition result of the text to be detected.

[0091] In some embodiments, the fraud mode includes: one or more of an identity impersonation fraud mode, a false investment fraud mode, and a lottery fraud mode.

[0092] In some embodiments, the emotion category includes one or more of fear emotion, tension emotion, anxiety emotion, and excitement emotion.

[0093] In some embodiments, the keywords include: one or more of: bank account, remittance, bonus, password.

[0094] In some embodiments, the device further comprises a model training module for:

[0095] Get historical fraud text data;

[0096] Classify the historical fraud text data to obtain a fraud pattern discrimination data set, a sentiment analysis data set, and a keyword extraction data set;

[0097] Training a first benchmark model based on the fraud pattern discrimination data set to obtain the fraud pattern discrimination model;

[0098] Training a second benchmark model based on the sentiment analysis data set to obtain the sentiment analysis model;

[0099] A third benchmark model is trained based on the keyword extraction data set to obtain the keyword extraction model.

[0100] In some embodiments, the device further comprises a data cleaning module, which is used to:

[0101] The historical fraud text data is cleaned; wherein the data cleaning step includes: removing duplicate data, correcting format problems in the text, filtering out one or more of preset irrelevant information.

[0102] In some embodiments, the apparatus further comprises a model updating module, configured to:

[0103] Obtain target fraud text data; the target fraud text data is fraud text data collected after a preset time interval of completing training of the fraud pattern discrimination model, the sentiment analysis model, and the keyword extraction model;

[0104] The parameters of the fraud pattern discrimination model, the sentiment analysis model and the keyword extraction model are updated based on the target fraud text data.

[0105] For the convenience of description, the above device is described in terms of functions divided into various modules. Of course, when implementing the present application, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0106] The fraud intention identification device of the above-mentioned embodiment is used to implement the corresponding fraud intention identification method in any of the above-mentioned embodiments, and has the beneficial effects of the corresponding fraud intention identification method embodiment, which will not be repeated here.

[0107] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for identifying fraudulent intent described in any of the above embodiments is implemented.

[0108] It should be noted that the method of the embodiment of the present application can be performed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present application, and the multiple devices will interact with each other to complete the described method.

[0109] It should be noted that the above describes some embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0110] Figure 4 A more specific schematic diagram of the hardware structure of an electronic device provided in this embodiment is shown. The device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040 and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030 and the communication interface 1040 are connected to each other in communication within the device through the bus 1050.

[0111] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0112] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0113] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0114] The communication interface 1040 is used to connect a communication module (not shown) to realize communication interaction between the device and the power equipment. The communication module can realize communication through a wired method (such as USB, network cable, etc.) or a wireless method (such as mobile network, WIFI, Bluetooth, etc.).

[0115] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0116] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.

[0117] The electronic device of the above embodiment is used to implement the corresponding fraud intention identification method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.

[0118] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present disclosure also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the fraud intention identification method described in any of the above embodiments.

[0119] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0120] The computer instructions stored in the storage medium of the above embodiments are used to enable the computer to execute the fraud intention identification method described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0121] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present application also provides a computer program product, which includes a computer program. In some embodiments, the computer program instructions can be executed by one or more processors of a computer so that the computer and / or the processor execute the fraud intention identification method described in the above-mentioned embodiments. Corresponding to the execution subject corresponding to each step in each embodiment of the fraud intention identification method, the processor that executes the corresponding step can belong to the corresponding execution subject.

[0122] To simplify the description and discussion, and in order not to make the embodiments of the present application difficult to understand, the known power supply / ground connection with the integrated circuit (IC) chip and other components may or may not be shown in the provided drawings. In addition, the device can be shown in the form of a block diagram to avoid making the embodiments of the present application difficult to understand, and this also takes into account the following fact, that is, the details of the implementation of these block diagram devices are highly dependent on the platform to be implemented in the embodiments of the present application (that is, these details should be fully within the scope of understanding of those skilled in the art). In the case of elaborating specific details (e.g., circuit) to describe exemplary embodiments of the present application, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details or when these specific details are changed. Therefore, these descriptions should be considered to be illustrative rather than restrictive.

[0123] Although the present application has been described in conjunction with specific embodiments of the present application, many replacements, modifications and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may use the embodiments discussed.

[0124] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present application (including the claims) is limited to these examples. In line with the concept of the present application, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the present application as described above, which are not provided in detail for the sake of simplicity.

[0125] The embodiments of the present application are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present application should be included in the scope of protection of the present application.

Claims

1. A method for identifying fraudulent intent, characterized in that: include: Get the text to be detected; The text to be detected is respectively input into the fraud pattern discrimination model, sentiment analysis model and keyword extraction model obtained through training to obtain the fraud pattern, sentiment category and keyword corresponding to the text to be detected; The text to be detected, the fraud pattern, the emotion category and the keyword are input into the fraud intention recognition model obtained through training to obtain the fraud intention recognition result of the text to be detected.

2. The method according to claim 1, characterized in that The fraud modes include: one or more of: impersonation fraud mode, false investment fraud mode, and lottery fraud mode.

3. The method according to claim 1, characterized in that The emotion categories include one or more of fear emotion, tension emotion, anxiety emotion, and excitement emotion.

4. The method according to claim 1, characterized in that: The keywords include: one or more of: bank account, remittance, bonus, and password.

5. The method according to claim 1, characterized in that Before inputting the to-be-detected text into the trained fraud pattern discrimination model, sentiment analysis model and keyword extraction model respectively, the method further comprises: Get historical fraud text data; Classify the historical fraud text data to obtain a fraud pattern discrimination data set, a sentiment analysis data set, and a keyword extraction data set; Training a first benchmark model based on the fraud pattern discrimination data set to obtain the fraud pattern discrimination model; Training a second benchmark model based on the sentiment analysis data set to obtain the sentiment analysis model; A third benchmark model is trained based on the keyword extraction data set to obtain the keyword extraction model.

6. The method according to claim 5, characterized in that Before classifying the historical fraud text data, the method further includes: The historical fraud text data is cleaned; wherein the data cleaning step includes: removing duplicate data, correcting format problems in the text, filtering out one or more of preset irrelevant information.

7. The method according to claim 1, characterized in that The method further comprises: Obtain target fraud text data; the target fraud text data is fraud text data collected after a preset time interval of completing training of the fraud pattern discrimination model, the sentiment analysis model, and the keyword extraction model; The parameters of the fraud pattern discrimination model, the sentiment analysis model and the keyword extraction model are updated based on the target fraud text data.

8. A fraudulent intention identification device, characterized in that: include: Get module to get the text to be detected; The prompt word module inputs the text to be detected into the fraud pattern discrimination model, sentiment analysis model and keyword extraction model obtained through training, respectively, to obtain the fraud pattern, sentiment category and keyword corresponding to the text to be detected; The recognition module inputs the text to be detected, the fraud pattern, the emotion category and the keyword into the fraud intention recognition model obtained through training to obtain the fraud intention recognition result of the text to be detected.

9. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 7.

Images