Risk early warning method and device, equipment, storage medium and program product

Abstract

The invention discloses a risk early warning method, device and equipment, a storage medium and a program product. The invention relates to the technical field of data processing. Acquiring news media data, and identifying entities and emotion information of the news media data; the method comprises the following steps: screening entities in news media data with negative emotion information based on a preset entity list to obtain a target entity; extracting a target text corresponding to the target entity from news media data where the target entity is located; and processing the target text based on a preset neural network model to identify a target entity of which the emotion information is negative and is an enterprise entity, and generating early warning information. According to the risk early warning method provided by the embodiment of the invention, the negative enterprise entity is extracted from the negative news media data to carry out risk early warning on the financial institution, so that the accuracy of negative enterprise entity identification can be improved, and the reliability of risk early warning can also be improved.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of data processing technology, and in particular, to a risk warning method, device, equipment, storage medium and program product. Background Art

[0002] With the rapid development of Internet technology, the speed of generating media data (such as news) is gradually increasing, so it is necessary to process massive data. In the financial field, financial news plays a very important role in assessing the credit risk of enterprises. By identifying the emotions of corporate entities in financial news, we can effectively understand the positive or negative reports of enterprises, thus providing a basis for further adjusting the credit risk rating of enterprises. The existing technology has low accuracy in identifying the emotions of corporate entities in news, which affects the reliability of risk warning. Summary of the invention

[0003] The embodiments of the present invention provide a risk warning method, apparatus, device, storage medium and program product, which can realize the identification of corporate entities with negative emotions in media data, thereby improving the reliability of risk warning.

[0004] In a first aspect, an embodiment of the present invention provides a risk warning method, comprising:

[0005] Acquire news media data, and identify entities and sentiment information of the news media data; wherein the sentiment information includes positive, negative and neutral;

[0006] Filter entities in news media data with negative sentiment information based on a preset entity list to obtain target entities;

[0007] Extracting a target text corresponding to the target entity from the news media data where the target entity is located; wherein the target text includes a subtext of the target entity located in the news media data and upper and lower subtexts of the subtext;

[0008] The target text is processed based on a preset neural network model to identify a target entity whose sentiment information is negative and is a corporate entity, and generate warning information.

[0009] In a second aspect, an embodiment of the present invention further provides a risk warning device, including:

[0010] An entity and sentiment information recognition module, used to obtain news media data and identify entities and sentiment information of the news media data; wherein the sentiment information includes positive, negative and neutral;

[0011] A target entity acquisition module is used to filter entities in news media data with negative sentiment information based on a preset entity list to obtain a target entity;

[0012] A target text extraction module is used to extract a target text corresponding to the target entity from the news media data where the target entity is located; wherein the target text includes a subtext of the target entity in the news media data and upper and lower subtexts of the subtext;

[0013] The target text processing module is used to process the target text based on a preset neural network model to identify the target entity whose sentiment information is negative and is a corporate entity, and generate early warning information.

[0014] In a third aspect, an embodiment of the present invention further provides an electronic device, comprising 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 risk warning method as described in the embodiment of the present invention is implemented.

[0015] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the risk warning method as described in the embodiment of the present invention is implemented.

[0016] In a fifth aspect, an embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the risk warning method as described in the embodiment of the present invention.

[0017] The embodiment of the present invention discloses a risk warning method, device, equipment, storage medium and program product, which obtain news media data and identify entities and sentiment information of the news media data; wherein the sentiment information includes positive, negative and neutral; based on a preset entity list, entities in the news media data with negative sentiment information are screened to obtain a target entity; a target text corresponding to the target entity is extracted from the news media data where the target entity is located; wherein the target text includes the subtext of the target entity in the news media data and the upper and lower subtexts of the subtext; the target text is processed based on a preset neural network model to identify the target entity with negative sentiment information and a corporate entity, and generate warning information. The risk warning method provided by the embodiment of the present invention extracts negative corporate entities from negative news media data to provide risk warnings to financial institutions, which can not only improve the accuracy of negative corporate entity identification, but also improve the reliability of risk warnings. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 A flow chart of a risk warning method provided by an embodiment of the present invention;

[0020] Figure 2 A flow chart of a risk warning method provided by an embodiment of the present invention;

[0021] Figure 3 A schematic diagram of the structure of a risk warning device provided by an embodiment of the present invention;

[0022] Figure 4 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0024] It should be noted that similar numbers and letters represent similar items in the following figures. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance. The acquisition, storage, use, processing, etc. of data in the technical solution of this application comply with the relevant provisions of national laws and regulations. It should be noted that in the embodiments of this application, certain software, components, models and other existing solutions in the industry may be mentioned, and they should be regarded as exemplary. The purpose is only to illustrate the feasibility of the implementation of the technical solution of this application, but it does not mean that the applicant has or will necessarily use the solution.

[0025] Figure 1 The present invention provides a flowchart of a risk warning method according to an embodiment of the present invention. The method can be applied to identifying negative corporate entities in negative news media data to provide risk warnings. The method can be executed by a risk warning device, which can be implemented in the form of software and / or hardware. Optionally, the device can be implemented by an electronic device, which can be a mobile terminal, a PC, or a server. Figure 1 As shown, the method includes:

[0026] S110, acquiring news media data, and identifying entities and sentiment information of the news media data.

[0027] The emotional information includes positive, negative and neutral. The news media data may be the latest news media data obtained from various news websites on the Internet. The news media data in this embodiment may be news media data in a news media source set by a financial institution.

[0028] Specifically, the method of obtaining news media data may be: obtaining original news media data and a news media source list; filtering original news daily data based on the news media source list to extract news media data belonging to news media sources in the news media source list.

[0029] Among them, the news media source list can be a list consisting of multiple news media sources that the financial institution pre-configures and is concerned about. The news media source can be understood as the source of news media data, such as a news network or a news platform. In this embodiment, after obtaining the original news media data, the source of the original news media data and the news media source are first determined, and then the news media source is compared with the news media source list. If the news media source falls into the news media source list, the news media data of the news media source is retained. If the news media source does not fall into the news media source list, the news media data of the news media source is deleted. In this embodiment, only the news media data corresponding to the news media source list is retained, which can not only greatly reduce the amount of data processing, but also help to improve the accuracy of negative corporate entity extraction.

[0030] Among them, the entity in the news media data can be reflected in the form of an entity name (such as a full name or abbreviation), and the entity can represent an institutional entity, an organizational entity or a corporate entity.

[0031] In this embodiment, the method of identifying entities and sentiment information of news media data can be: inputting news media data into an entity recognition model and outputting entities; inputting news media data into a sentiment recognition model and outputting sentiment information.

[0032] Among them, both the entity recognition model and the emotion recognition model are neural network models pre-trained based on a large number of samples. This embodiment does not limit the network structure and training method of the entity recognition model and the emotion recognition model. The trained entity recognition model has the ability to recognize entities in news media data, and the emotion recognition model has the ability to recognize emotional information of news media data. In this embodiment, after obtaining the news media data, the news media data can first be pre-processed by cleaning, standardization, etc., and then the pre-processed news media data can be input into the entity recognition model to output the entity; the pre-processed news media data can be input into the emotion recognition model to output the emotional information. In this embodiment, the recognition accuracy and efficiency can be improved by using the pre-trained neural network model to recognize entities in news media data and emotional information of news media data.

[0033] Optionally, after the news media data is input into the entity recognition model and the entities are output, the following step is also included: filtering the output entities based on a preset blacklist and / or regular expression to obtain filtered entities.

[0034] The preset blacklist records the organizational entities that the financial institution does not pay attention to. The regular expression can be a preset expression for filtering the output entities. For example, the regular expression in this embodiment can filter out entities containing preset text, and the preset text can be: Research, Market, StockExchange, Commission, Department.

[0035] In this embodiment, since financial institutions only focus on corporate entities, and the entities identified by the entity recognition model can be institutional entities, organizational entities or corporate entities, it is necessary to further screen the identified entities. Specifically, the method of filtering the output entities based on a preset blacklist and / or regular expression can be: if the output entity falls into the preset blacklist and / or satisfies the regular expression, the entity is deleted to obtain the filtered entity. In this embodiment, filtering the output entities based on a preset blacklist and / or regular expression can further reduce the amount of calculation and improve the efficiency of identifying negative corporate entities.

[0036] S120, filtering entities in news media data with negative sentiment information based on a preset entity list to obtain a target entity.

[0037] The preset entity list may be a list of enterprise names that are pre-configured by a financial institution and that the financial institution is concerned about, and the enterprise name may include the full name and / or the abbreviation of the enterprise. The preset entity list may be generated by: obtaining the full name of the entity; extracting the abbreviation of the entity from the full name of the entity; and generating the preset entity list based on the full name of the entity and / or the abbreviation of the entity.

[0038] The full name of the entity can be the full name in Chinese or in English. If the full name of the entity is in English, the abbreviation of the entity can be extracted from the full name by removing the suffix in the full name of the entity or extracting the first letter of the full English name of the enterprise.

[0039] The suffix in the full name of the entity may be the suffix added when the English full name of the entity is named. Exemplary, the suffix may be: LTD, SA, LIMITED, CO.LTD., PLC, INC., AG, etc. The full English name includes multiple English words, and the first letter of each English word is extracted, and the extracted first letter is used as the abbreviation of the entity. If the full name of the entity is a Chinese full name, the method of extracting the entity abbreviation from the full name of the entity may be: extracting keywords from the Chinese full name as the abbreviation of the entity.

[0040] In this embodiment, the process of screening entities in news media data with negative sentiment information based on a preset entity list may be: first extracting entities from the negative news media data, and then screening out entities that fall into the preset entity list.

[0041] Specifically, entities in news media data with negative sentiment information are screened based on a preset entity list, and a method for obtaining a target entity may be: entities in news media data with negative sentiment information that match entities in the preset entity list are screened out to obtain the target entity.

[0042] Among them, the entities in the preset entity list are represented by names, and the names include full names and / or abbreviations. In this embodiment, for each entity in the negative news media data, the entity is compared with the entity abbreviation or full name in the preset entity list. If the entity abbreviation or full name in the preset entity list matches, the entity is determined as the target entity; if the entity abbreviation or full name in the preset entity list matches, the entity is discarded. In this embodiment, based on the preset entity list, entities in news media data with negative sentiment information are screened, and entities of concern to financial institutions can be screened out, thereby improving the accuracy of risk warnings.

[0043] S130, extracting a target text corresponding to the target entity from the news media data where the target entity is located.

[0044] Among them, the target text includes the subtext where the target entity is located in the news media data and the upper and lower subtexts of the subtext. The subtext where the target entity is located in the news media data can be understood as the sentence where the target entity is located in the news media data; the upper and lower subtexts of the subtext can be the upper N sentences and lower N sentences of the sentence where the target entity is located. The sentence where the target entity is located in the news media data, the upper N sentences and the lower N sentences constitute the target text corresponding to the target entity. Among them, N is an integer greater than or equal to 1. Exemplarily, assuming that N is 2, a target entity is in the 50th sentence of the news media data, the upper subtext is the 48th and 49th sentences, and the lower subtext is the 51st and 52nd sentences, then the target text corresponding to the target entity is composed of the 48th to 49th sentences of the news media data where it is located.

[0045] S140, processing the target text based on a preset neural network model to identify the target entity whose sentiment information is negative and is a corporate entity, and generating warning information.

[0046] The preset neural network model may be a pre-trained large language model (LLM). In this application scenario, when using the preset entity list to screen entities in news media data with negative sentiment information, since the entity abbreviations in the list are used, different entities may have the same abbreviations, so the screened target entity may be a non-enterprise entity, so it is necessary to use the preset neural network model to further identify the target entity.

[0047] Specifically, the process of processing the target text based on the preset neural network model to identify the target entity whose sentiment information is negative and which is a corporate entity can be: generating a first prompt message; inputting the first prompt message, the target entity and the target text into the preset neural network model, and outputting the entity recognition result; if the entity recognition result is that the target entity is a corporate entity, generating a second prompt message, and inputting the second prompt message, the target entity, the entity recognition result and the target text into the preset neural network model, and outputting the sentiment information of the target entity.

[0048] Among them, the first prompt information is used to prompt the preset neural network model to identify whether the target entity is a corporate entity, and the second prompt information is used to prompt the preset neural network model to identify the emotional information of the target entity. The entity recognition result includes whether the target entity is a corporate entity or the target entity is a non-corporate entity. The emotional information of the target entity can include: positive, negative and neutral. The first prompt information and the second prompt information are both edited and generated by the user according to actual needs. For example: assuming that the target entity is A, the first prompt information can be "Please determine whether A is a corporate entity based on the following content", and the second prompt information can be "Please confirm the emotional information of corporate entity A"

[0049] In this embodiment, after the first prompt information is generated, the first prompt information, the target entity and the target text are input into a preset neural network model, so that the preset neural network model identifies the target entity in the target text based on the first prompt information. If the target entity is an enterprise entity, the entity recognition result that the target entity is an enterprise entity is output; if the target entity is a non-enterprise entity, the entity recognition result that the target entity is a non-enterprise entity is output.

[0050] Specifically, if the target entity is a non-enterprise entity, the target entity is discarded; if the target entity is an enterprise entity, the second prompt information, the target entity, the entity recognition result and the target text continue to be input into the preset neural network model, so that the preset neural network model recognizes the emotional information of the target entity based on the second prompt information, and outputs the recognized emotional information.

[0051] In this embodiment, if the target entity is a corporate entity and the sentiment information is negative, a warning message is generated. The warning message can be generated using a preset warning message generation module or a preset neural network model.

[0052] Specifically, the method for generating warning information can be: if the target entity is a corporate entity and the emotional information is negative, a third prompt information is generated; the third prompt information, entity recognition results, emotional information and news media data where the target entity is located are input into a preset neural network model, and a warning information is output.

[0053] Among them, the third prompt information is used to prompt the preset neural network model to generate warning information; the warning information includes: warning reasons and risk points. The third prompt information is edited and generated by the user according to actual needs. For example, assuming that the target entity is A, the third prompt information can be "generate warning information about A that includes warning reasons and risk points." In this embodiment, after the third prompt information is generated, the third prompt information, the entity recognition result of the target entity being a corporate entity, the emotional information being negative, and the news media data where the target entity is located are input into the preset neural network model, so that the preset neural network model extracts the warning reasons and risk points from the news media data where the target entity is located based on the third prompt information, and outputs the warning reasons and risk points.

[0054] Optionally, after the warning information is generated, the following steps are also included: displaying the warning information for the target user to review; in response to the review passing operation triggered by the target user, pushing the warning information to the target terminal.

[0055] In this embodiment, the warning information is displayed on the front end so that the target user can manually review the warning information. If the manual review is passed, the target user clicks the review pass control, and the warning information is sent to the target terminal for further analysis and review. Manual review of the warning information can improve the accuracy of the warning information.

[0056] The technical solution of this embodiment obtains news media data and identifies entities and sentiment information of the news media data; wherein the sentiment information includes positive, negative and neutral; based on a preset entity list, entities in the news media data with negative sentiment information are screened to obtain a target entity; a target text corresponding to the target entity is extracted from the news media data where the target entity is located; wherein the target text includes the subtext of the target entity in the news media data and the upper and lower subtexts of the subtext; the target text is processed based on a preset neural network model to identify the target entity with negative sentiment information and a corporate entity, and generate warning information. The risk warning method provided by the embodiment of the present invention extracts negative corporate entities from negative news media data to provide risk warnings to financial institutions, which can not only improve the accuracy of negative corporate entity identification, but also improve the reliability of risk warnings.

[0057] Based on the above embodiments, Figure 2 is a flow chart of a risk warning method provided by an embodiment of the present invention, such as Figure 2 As shown, the method includes:

[0058] S210, obtaining original news media data and a news media source list; filtering the original news data based on the news media source list to extract news media data belonging to the news media sources in the news media source list

[0059] S220, input the news media data into the entity recognition model and output the entity; input the news media data into the emotion recognition model and output the emotion information.

[0060] S230, filtering the output entities based on a preset blacklist and / or regular expression to obtain filtered entities.

[0061] S240, filtering out entities in news media data with negative sentiment information that match entities in a preset entity list to obtain a target entity.

[0062] S250, extracting a target text corresponding to the target entity from the news media data where the target entity is located.

[0063] S260, generating first prompt information, inputting the first prompt information, the target entity and the target text into a preset neural network model, and outputting an entity recognition result.

[0064] S270, if the entity recognition result shows that the target entity is a corporate entity, a second prompt message is generated, and the second prompt message, the target entity, the entity recognition result and the target text are input into a preset neural network model to output the emotional information of the target entity.

[0065] S280, if the target entity is a corporate entity and the emotional information is negative, a third prompt message is generated, and the third prompt message, entity recognition result, emotional information and news media data where the target entity is located are input into a preset neural network model to output a warning message.

[0066] S290, displaying the warning information for the target user to review; in response to the review passing operation triggered by the target user, pushing the warning information to the target terminal.

[0067] Figure 3 is a schematic diagram of the structure of a risk warning device provided by an embodiment of the present invention, such as Figure 3 As shown, the device comprises:

[0068] The entity and sentiment information identification module 310 is used to obtain news media data and identify entities and sentiment information of the news media data; wherein the sentiment information includes positive, negative and neutral;

[0069] A target entity acquisition module 320 is used to filter entities in news media data with negative sentiment information based on a preset entity list to obtain a target entity;

[0070] The target text extraction module 330 is used to extract the target text corresponding to the target entity from the news media data where the target entity is located; wherein the target text includes the subtext where the target entity is located in the news media data and the upper and lower subtexts of the subtext;

[0071] The target text processing module 340 is used to process the target text based on a preset neural network model to identify the target entity whose sentiment information is negative and is a corporate entity, and generate warning information.

[0072] Optionally, the entity and sentiment information identification module 310 is further used to:

[0073] Obtain raw news media data and a list of news media sources;

[0074] The original news daily data is screened based on the news media source list to extract the news media data belonging to the news media sources in the news media source list.

[0075] Optionally, the entity and sentiment information identification module 310 is further used to:

[0076] Input news media data into the entity recognition model and output entities;

[0077] Input news media data into the sentiment recognition model and output sentiment information.

[0078] Optionally, it also includes: an entity filtering module, which is used to:

[0079] The output entities are filtered based on a preset blacklist and / or regular expression to obtain filtered entities.

[0080] Optionally, the target entity acquisition module 320 is further configured to:

[0081] Entities matching entities in a preset entity list are screened out from news media data with negative sentiment information to obtain target entities; wherein the entities in the preset entity list are represented by names, and the names include full names and / or abbreviations.

[0082] Optionally, the target text processing module 340 is further used to:

[0083] Generate first prompt information; wherein the first prompt information is used to prompt a preset neural network model to identify whether the target entity is a corporate entity;

[0084] Inputting the first prompt information, the target entity and the target text into a preset neural network model, and outputting an entity recognition result; wherein the entity recognition result includes whether the target entity is an enterprise entity or the target entity is a non-enterprise entity;

[0085] If the entity recognition result shows that the target entity is a corporate entity, a second prompt information is generated, and the second prompt information, the target entity, the entity recognition result and the target text are input into a preset neural network model to output the emotional information of the target entity; wherein the second prompt information is used to prompt the preset neural network model to recognize the emotional information of the target entity.

[0086] Optionally, the target text processing module 340 is further used to:

[0087] If the target entity is a corporate entity and the sentiment information is negative, a third prompt information is generated; wherein the third prompt information is used to prompt the preset neural network model to generate warning information;

[0088] The third prompt information, entity recognition results, emotional information and news media data where the target entity is located are input into a preset neural network model, and warning information is output; wherein the warning information includes: warning reasons and risk points.

[0089] Optionally, it also includes: an early warning information push module, which is used to:

[0090] Display the warning information for target users to review;

[0091] In response to the audit passing operation triggered by the target user, the warning information is pushed to the target terminal.

[0092] Optionally, it also includes: a preset entity list generation module, which is used to:

[0093] Get the full name of the entity;

[0094] Extract entity abbreviation from entity full name;

[0095] Generate a preset entity list based on the full entity name and / or the short entity name.

[0096] Optionally, the preset entity list generation module is also used to:

[0097] Remove the suffix from the entity's full name; or,

[0098] Extract the first letter of the full English name of the company.

[0099] The above device can execute the methods provided by all the above embodiments of the present invention, and has the corresponding functional modules and beneficial effects of executing the above methods. For technical details not described in detail in this embodiment, please refer to the methods provided by all the above embodiments of the present invention.

[0100] Figure 4 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0101] like Figure 4As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0102] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0103] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the risk warning method.

[0104] In some embodiments, the risk warning method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the risk warning method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute the risk warning method in any other appropriate manner (e.g., by means of firmware).

[0105] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0106] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0107] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0108] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0109] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0110] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0111] An embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the risk warning method provided in any embodiment of the present application.

[0112] In the process of implementation, the computer program product can be written in one or more programming languages ​​or a combination thereof to perform the computer program code of the present invention, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0113] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A risk early warning method, characterized in that: include: Acquire news media data, and identify entities and sentiment information of the news media data; wherein the sentiment information includes positive, negative and neutral; Filter entities in news media data with negative sentiment information based on a preset entity list to obtain target entities; Extracting a target text corresponding to the target entity from the news media data where the target entity is located; wherein the target text includes a subtext of the target entity located in the news media data and upper and lower subtexts of the subtext; The target text is processed based on a preset neural network model to identify a target entity whose sentiment information is negative and is a corporate entity, and generate warning information.

2. The method according to claim 1, characterized in that Access news media data, including: Obtain raw news media data and a list of news media sources; The original news daily data is screened based on the news media source list to extract news media data belonging to the news media sources in the news media source list.

3. The method according to claim 1, characterized in that Identify entities and sentiment information of the news media data, including: Input the news media data into an entity recognition model and output entities; The news media data is input into a sentiment recognition model to output sentiment information.

4. The method according to claim 3, characterized in that: After the news media data is input into the entity recognition model and the entity is output, the method further includes: The output entities are filtered based on a preset blacklist and / or regular expression to obtain filtered entities.

5. The method according to claim 1, characterized in that Based on the preset entity list, the entities in the news media data with negative sentiment information are screened to obtain the target entities, including: Entities matching entities in a preset entity list are screened out from news media data with negative sentiment information to obtain target entities; wherein the entities in the preset entity list are represented by names, and the names include full names and / or abbreviations.

6. The method according to claim 1, characterized in that The target text is processed based on a preset neural network model to identify a target entity having negative sentiment information and being a corporate entity, including: Generate first prompt information; wherein the first prompt information is used to prompt the preset neural network model to identify whether the target entity is a corporate entity; Inputting the first prompt information, the target entity and the target text into the preset neural network model, and outputting an entity recognition result; wherein the entity recognition result includes whether the target entity is an enterprise entity or the target entity is a non-enterprise entity; If the entity recognition result shows that the target entity is a corporate entity, a second prompt information is generated, and the second prompt information, the target entity, the entity recognition result and the target text are input into the preset neural network model to output the emotional information of the target entity; wherein the second prompt information is used to prompt the preset neural network model to recognize the emotional information of the target entity.

7. The method according to claim 6, characterized in that Generate early warning information, including: If the target entity is a corporate entity and the sentiment information is negative, a third prompt information is generated; wherein the third prompt information is used to prompt the preset neural network model to generate warning information; The third prompt information, the entity recognition result, the emotional information and the news media data where the target entity is located are input into the preset neural network model, and warning information is output; wherein the warning information includes: warning reasons and risk points.

8. The method according to claim 1, characterized in that After the warning information is generated, it also includes: Displaying the warning information for review by target users; In response to the review passing operation triggered by the target user, the warning information is pushed to the target terminal.

9. The method according to claim 5, characterized in that The default entity list is generated as follows: Get the full name of the entity; Extracting an entity abbreviation from the entity full name; A preset entity list is generated based on the full name of the entity and / or the abbreviation of the entity.

10. The method according to claim 9, characterized in that Extracting the entity abbreviation from the entity full name, including: Remove the suffix from the entity's full name; or, Extract the first letters of the full English name of the company.

11. A risk warning device, characterized in that: include: An entity and sentiment information recognition module, used to obtain news media data and identify entities and sentiment information of the news media data; wherein the sentiment information includes positive, negative and neutral; A target entity acquisition module is used to filter entities in news media data with negative sentiment information based on a preset entity list to obtain a target entity; A target text extraction module is used to extract a target text corresponding to the target entity from the news media data where the target entity is located; wherein the target text includes a subtext of the target entity in the news media data and upper and lower subtexts of the subtext; The target text processing module is used to process the target text based on a preset neural network model to identify the target entity whose sentiment information is negative and is a corporate entity, and generate early warning information.

12. The device according to claim 11, characterized in that The entity and sentiment information recognition module is also used for: Input the news media data into an entity recognition model and output entities; The news media data is input into a sentiment recognition model to output sentiment information.

13. 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 the processor implements the risk warning method according to any one of claims 1 to 10 when executing the computer program.

14. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the risk warning method as described in any one of claims 1 to 10 is implemented.

15. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the computer program implements the risk warning method according to any one of claims 1 to 10.

Images