Information processing method and device, equipment, storage medium and computer program product

Abstract

The invention provides an information processing method and device, equipment, a storage medium and a computer program product.The information processing method comprises the steps that identification information of a target object is determined; according to the identification information, obtaining analysis reference information corresponding to the target object; the analysis reference information comprises industry information, market feedback information, historical trend information and technical index information; and determining an analysis result corresponding to the target object according to the analysis reference information. According to the scheme, analysis can be carried out according to more comprehensive information, comprehensive consideration is carried out on related information, the relevance between the information is ensured, comprehensive judgment is formed, a comprehensive and professional analysis result is provided, the analysis accuracy is improved, and the analysis efficiency is improved. The problem that in the prior art, related schemes for data analysis are low in analysis accuracy is well solved.

Description

Technical Field

[0001] The present application relates to the field of information processing technology, and in particular to an information processing method, apparatus, device, storage medium and computer program product. Background Art

[0002] With the continuous development of the financial market and the increasing demand of investors for analysis of individual stocks (such as funds or industries), traditional analysis methods can no longer meet the requirements of multi-dimensional and in-depth analysis. The existing solutions have the following defects:

[0003] (1) Incomplete information: Traditional stock analysis methods mainly rely on human analysts to conduct analysis and prediction by studying fundamental, technical and other data. However, it is difficult to fully capture market dynamics, resulting in incomplete analysis and low accuracy.

[0004] (2) Information is not correlated: Traditional analysis methods often lack effective integration methods when processing information from different sources, resulting in weak correlation between information, making it difficult to form a comprehensive judgment, and resulting in low analysis accuracy.

[0005] As described above, the related solutions for data analysis in the prior art have problems such as low analysis accuracy. Summary of the invention

[0006] The purpose of the present application is to provide an information processing method, apparatus, device, storage medium and computer program product to solve the problem of low analysis accuracy in related solutions for data analysis in the prior art.

[0007] In order to solve the above technical problems, the present application provides an information processing method, including:

[0008] Determine identification information of the target object;

[0009] According to the identification information, obtaining analysis reference information corresponding to the target object; the analysis reference information includes: industry information, market feedback information, historical trend information and technical indicator information;

[0010] An analysis result corresponding to the target object is determined according to the analysis reference information.

[0011] Optionally, determining the analysis result corresponding to the target object according to the analysis reference information includes:

[0012] Using the first agent, according to the industry information in the analysis reference information, output information is obtained, wherein the output information includes: industry status information and prospect information corresponding to the target object;

[0013] Using a second agent, based on the market feedback information in the analysis reference information, assess the degree of influence of the market feedback on the target object, and obtain first prediction information corresponding to the target object;

[0014] Using a third agent, quantitatively analyzing the historical trend information and the technical indicator information in the analysis reference information to obtain second prediction information corresponding to the target object;

[0015] An analysis result corresponding to the target object is obtained according to the output information, the first prediction information and the second prediction information.

[0016] Optionally, the using the first agent to obtain output information according to the industry information in the analysis reference information includes:

[0017] Using a first agent, output information is obtained according to first information and industry information in the analysis reference information; the first information includes: the first prediction information and a first weight corresponding to the first prediction information; and / or the second prediction information and a second weight corresponding to the second prediction information;

[0018] And / or, the using of the second agent to evaluate the degree of influence of market feedback on the target object according to the market feedback information in the analysis reference information to obtain the first prediction information corresponding to the target object includes:

[0019] Using the second agent, based on the second information and the market feedback information in the analysis reference information, the degree of influence of the market feedback on the target object is evaluated to obtain the first prediction information corresponding to the target object; the second information includes: the output information and the third weight corresponding to the output information; and / or, the second prediction information and the second weight corresponding to the second prediction information;

[0020] And / or, the using of a third agent to perform quantitative analysis on the historical trend information and technical indicator information in the analysis reference information to obtain second prediction information corresponding to the target object includes:

[0021] Using a third agent, based on the third information, a quantitative analysis is performed on the historical trend information and technical indicator information in the analysis reference information to obtain second prediction information corresponding to the target object; the second information includes: the output information and the third weight corresponding to the output information; and / or, the first prediction information and the first weight corresponding to the first prediction information.

[0022] Optionally, obtaining an analysis result corresponding to the target object according to the output information, the first prediction information, and the second prediction information includes:

[0023] Determining, using a fourth agent, whether the output information, the first prediction information, and the second prediction information match;

[0024] In the case of a match, an analysis result corresponding to the target object is obtained according to the output information, the first prediction information, the second prediction information and the user characteristic information.

[0025] Optionally, also include:

[0026] In the case where the output information, the first prediction information, and the second prediction information do not match, obtaining a user opinion;

[0027] Obtaining an analysis result corresponding to the target object according to the output information, the first prediction information, and the second prediction information includes:

[0028] An analysis result corresponding to the target object is obtained according to the output information, the first prediction information, the second prediction information, the user opinion and the user characteristic information.

[0029] Optionally, the identification information includes: an object name and an object index;

[0030] The acquiring, according to the identification information, analysis reference information corresponding to the target object includes:

[0031] According to the object name, query the information text from the database to obtain the industry information and user feedback related text;

[0032] Obtaining the market feedback information according to the user feedback related text;

[0033] According to the object index, the service interface is queried to obtain the historical trend information and technical indicator information.

[0034] The present application also provides an information processing device, including:

[0035] A first determination module, used to determine identification information of a target object;

[0036] A first acquisition module is used to acquire analysis reference information corresponding to the target object according to the identification information; the analysis reference information includes: industry information, market feedback information, historical trend information and technical indicator information;

[0037] The second determination module is used to determine the analysis result corresponding to the target object according to the analysis reference information.

[0038] Optionally, determining the analysis result corresponding to the target object according to the analysis reference information includes:

[0039] Using the first agent, according to the industry information in the analysis reference information, output information is obtained, wherein the output information includes: industry status information and prospect information corresponding to the target object;

[0040] Using a second agent, based on the market feedback information in the analysis reference information, assess the degree of influence of the market feedback on the target object, and obtain first prediction information corresponding to the target object;

[0041] Using a third agent, quantitatively analyzing the historical trend information and the technical indicator information in the analysis reference information to obtain second prediction information corresponding to the target object;

[0042] An analysis result corresponding to the target object is obtained according to the output information, the first prediction information and the second prediction information.

[0043] Optionally, the using the first agent to obtain output information according to the industry information in the analysis reference information includes:

[0044] Using a first agent, output information is obtained according to first information and industry information in the analysis reference information; the first information includes: the first prediction information and a first weight corresponding to the first prediction information; and / or the second prediction information and a second weight corresponding to the second prediction information;

[0045] And / or, the using of the second agent to evaluate the degree of influence of market feedback on the target object according to the market feedback information in the analysis reference information to obtain the first prediction information corresponding to the target object includes:

[0046] Using the second agent, based on the second information and the market feedback information in the analysis reference information, the degree of influence of the market feedback on the target object is evaluated to obtain the first prediction information corresponding to the target object; the second information includes: the output information and the third weight corresponding to the output information; and / or, the second prediction information and the second weight corresponding to the second prediction information;

[0047] And / or, the using of a third agent to perform quantitative analysis on the historical trend information and technical indicator information in the analysis reference information to obtain second prediction information corresponding to the target object includes:

[0048] Using a third agent, based on the third information, a quantitative analysis is performed on the historical trend information and technical indicator information in the analysis reference information to obtain second prediction information corresponding to the target object; the second information includes: the output information and the third weight corresponding to the output information; and / or, the first prediction information and the first weight corresponding to the first prediction information.

[0049] Optionally, obtaining an analysis result corresponding to the target object according to the output information, the first prediction information, and the second prediction information includes:

[0050] Determining, using a fourth agent, whether the output information, the first prediction information, and the second prediction information match;

[0051] In the case of a match, an analysis result corresponding to the target object is obtained according to the output information, the first prediction information, the second prediction information and the user characteristic information.

[0052] Optionally, also include:

[0053] A second acquisition module, used for acquiring user opinions when the output information, the first prediction information and the second prediction information do not match;

[0054] Obtaining an analysis result corresponding to the target object according to the output information, the first prediction information, and the second prediction information includes:

[0055] An analysis result corresponding to the target object is obtained according to the output information, the first prediction information, the second prediction information, the user opinion and the user characteristic information.

[0056] Optionally, the identification information includes: an object name and an object index;

[0057] The acquiring, according to the identification information, analysis reference information corresponding to the target object includes:

[0058] According to the object name, query the information text from the database to obtain the industry information and user feedback related text;

[0059] Obtaining the market feedback information according to the user feedback related text;

[0060] According to the object index, the service interface is queried to obtain the historical trend information and technical indicator information.

[0061] An embodiment of the present application also provides an information processing device, including a memory, a processor, and a program stored in the memory and executable on the processor; when the processor executes the program, the above-mentioned information processing method is implemented.

[0062] An embodiment of the present application also provides a readable storage medium on which a program is stored, and when the program is executed by a processor, the steps in the above-mentioned information processing method are implemented.

[0063] An embodiment of the present application also provides a computer program product, including computer instructions, which implement the steps of the above-mentioned information processing method when executed by a processor.

[0064] The beneficial effects of the above technical solution of the present application are as follows:

[0065] In the above scheme, the information processing method determines the identification information of the target object; based on the identification information, obtains the analysis reference information corresponding to the target object; the analysis reference information includes: industry information, market feedback information, historical trend information and technical indicator information; based on the analysis reference information, determines the analysis result corresponding to the target object; can support the analysis based on more comprehensive information, and comprehensively consider the relevant information to ensure the correlation between the information in order to form a comprehensive judgment, thereby providing comprehensive and professional analysis results, improving the analysis accuracy, and well solving the problem of low analysis accuracy in related schemes for data analysis in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 This is a flowchart of the information processing method according to an embodiment of the present application;

[0067] Figure 2 A schematic diagram of a specific implementation flow of the information processing method of an embodiment of the present application;

[0068] Figure 3 This is a schematic diagram of the functional framework of an embodiment of the present application;

[0069] Figure 4 This is a schematic diagram of the structure of an information processing device according to an embodiment of the present application. DETAILED DESCRIPTION

[0070] In order to make the technical problems, technical solutions and advantages to be solved by the present application clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0071] This application aims to solve the problem of low accuracy of information processing solutions for chart questions and answers in the existing technology, and provides an information processing method, such as Figure 1 As shown, including:

[0072] Step 11: Determine the identification information of the target object;

[0073] Step 12: According to the identification information, obtain the analysis reference information corresponding to the target object; the analysis reference information includes: industry information, market feedback information, historical trend information and technical indicator information;

[0074] Step 13: Determine the analysis result corresponding to the target object according to the analysis reference information.

[0075] Among them, target objects such as target stocks, target funds, target industries, etc. are not limited here.

[0076] The information processing method provided in the embodiment of the present application determines the identification information of the target object; based on the identification information, obtains the analysis reference information corresponding to the target object; the analysis reference information includes: industry information, market feedback information, historical trend information and technical indicator information; based on the analysis reference information, determines the analysis result corresponding to the target object; can support the implementation of analysis based on more comprehensive information, and comprehensively consider the relevant information to ensure the correlation between the information in order to form a comprehensive judgment, thereby providing comprehensive and professional analysis results, improving the analysis accuracy, and well solving the problem of low analysis accuracy in related solutions for data analysis in the prior art.

[0077] Wherein, the determination of the analysis result corresponding to the target object according to the analysis reference information includes: using a first agent to obtain output information according to the industry information in the analysis reference information, and the output information includes: industry status information and prospect information corresponding to the target object; using a second agent to evaluate the degree of influence of market feedback on the target object according to the market feedback information in the analysis reference information, and obtain the first prediction information corresponding to the target object; using a third agent to perform quantitative analysis on the historical trend information and technical indicator information in the analysis reference information, and obtain the second prediction information corresponding to the target object; according to the output information, the first prediction information and the second prediction information, the analysis result corresponding to the target object is obtained. In this way, the analysis result corresponding to the target object can be obtained accurately and intelligently. Wherein, the first prediction information can be the first trend prediction information, such as including long-term trend prediction information and / or short-term trend prediction information; and / or, the second prediction information can include long-term trend prediction information and short-term trend prediction information; but it is not limited to this. Further, the method can also include: using prompt words to construct at least one of the first agent, the second agent, the third agent and the fourth agent, which is not limited here.

[0078] In an embodiment of the present application, (1) the use of the first agent to obtain output information based on the industry information in the analysis reference information includes: using the first agent to obtain output information based on the first information and the industry information in the analysis reference information; the first information includes: the first prediction information and the first weight corresponding to the first prediction information; and / or, the second prediction information and the second weight corresponding to the second prediction information; and / or, (2) the use of the second agent to evaluate the impact of market feedback on the target object based on the market feedback information in the analysis reference information to obtain the first prediction information corresponding to the target object includes: using the second agent to evaluate the impact of market feedback on the target object based on the second information and the market feedback information in the analysis reference information degree, obtain the first prediction information corresponding to the target object; the second information includes: the output information and the third weight corresponding to the output information; and / or, the second prediction information and the second weight corresponding to the second prediction information; and / or, (3) the use of the third agent to perform quantitative analysis on the historical trend information and technical indicator information in the analysis reference information to obtain the second prediction information corresponding to the target object, including: using the third agent, according to the third information, to perform quantitative analysis on the historical trend information and technical indicator information in the analysis reference information to obtain the second prediction information corresponding to the target object; the second information includes: the output information and the third weight corresponding to the output information; and / or, the first prediction information and the first weight corresponding to the first prediction information. This can support the interaction between agents, and use the output information of one agent as the input of another agent, such as splicing it into the prompt of the next agent, which is not limited here.

[0079] Wherein, the analysis result corresponding to the target object is obtained according to the output information, the first prediction information and the second prediction information, including: using a fourth agent to determine whether the output information, the first prediction information and the second prediction information match; if they match, the analysis result corresponding to the target object is obtained according to the output information, the first prediction information and the second prediction information and the user characteristic information. In this way, the analysis result corresponding to the target object can be obtained more accurately. Wherein, the user characteristic information may include at least one of the following information: the amount of funds, the maximum tolerance, the object preference (such as the preference for individual stocks), etc.; this is not limited here. In addition, the operation of "obtaining the analysis result corresponding to the target object according to the output information, the first prediction information and the second prediction information and the user characteristic information" can be specifically executed by calling a large financial model, etc., but is not limited to this.

[0080] In the embodiment of the present application, the fourth agent can obtain the analysis result by combining the weights of the three agents; correspondingly, obtaining the analysis result corresponding to the target object according to the output information, the first prediction information, and the second prediction information may include: obtaining the analysis result corresponding to the target object according to the output information, the third weight corresponding to the output information, the first prediction information, the first weight corresponding to the first prediction information, the second prediction information, and the second weight corresponding to the second prediction information (this operation may also be in an and / or relationship with (1), (2), and (3) related to the first information and other information involved above); more specifically, obtaining the analysis result corresponding to the target object according to the output information, the first prediction information, the second prediction information, and the user feature information may include: obtaining the analysis result corresponding to the target object according to the output information, the third weight corresponding to the output information, the first prediction information, the first weight corresponding to the first prediction information, the second prediction information, the second weight corresponding to the second prediction information, and the user feature information; this can further improve the accuracy of the analysis result;

[0081] And / or, the fourth agent can combine the weights of the three agents to "determine whether the output information, the first prediction information, and the second prediction information match"; correspondingly, using the fourth agent to determine whether the output information, the first prediction information, and the second prediction information match can include: using the fourth agent to determine whether the output information, the first prediction information, and the second prediction information match according to the third weight corresponding to the output information, the first weight corresponding to the first prediction information, and the second weight corresponding to the second prediction information; this can improve the accuracy of the matching results.

[0082] Further, the information processing method further includes: obtaining user opinions when the output information, the first prediction information and the second prediction information do not match; the analysis result corresponding to the target object is obtained according to the output information, the first prediction information and the second prediction information, including: obtaining the analysis result corresponding to the target object according to the output information, the first prediction information, the second prediction information, the user opinions and the user characteristic information. In this way, the degree of fit with the actual needs of the user can be guaranteed to a certain extent. Among them, the user opinions can be input by the user, or obtained by screening existing opinions and opinions, which is not limited here; in addition, the analysis result corresponding to the target object is obtained according to the output information, the first prediction information, the second prediction information, the user opinions and the user characteristic information, which can include: obtaining the analysis result corresponding to the target object according to the output information, the third weight corresponding to the output information, the first prediction information, the first weight corresponding to the first prediction information, the second prediction information, the second weight corresponding to the second prediction information, the user opinions and the user characteristic information; in this way, the accuracy of the analysis result can be further improved. In addition, the operation of "obtaining analysis results corresponding to the target object based on the output information, the first prediction information, the second prediction information, the user opinions and the user characteristic information" can be specifically executed by calling a large financial model, etc., but is not limited to this.

[0083] Wherein, the identification information includes: object name and object index; the acquisition of analysis reference information corresponding to the target object according to the identification information includes: querying information text from a database according to the object name to obtain the industry information and user feedback related text; obtaining the market feedback information according to the user feedback related text; querying a service interface according to the object index to obtain the historical trend information and technical indicator information. In this way, each piece of information contained in the analysis reference information can be accurately acquired. Wherein, the object index may include an object code (such as a stock code), but is not limited thereto.

[0084] The information processing method provided in the embodiment of the present application is illustrated below by way of example, wherein the target object takes the target stock (i.e., the target stock) as an example, the object name takes the stock name as an example, and the object index takes the stock code as an example.

[0085] In view of the above technical problems, and taking into account: the low user participation in the prior art: the traditional analysis method often ignores the user's point of view, resulting in the analysis results being out of touch with the actual needs of the user; the embodiment of the present application provides an information processing method, which can be specifically implemented as a processing and analysis method for multi-source heterogeneous stock data, mainly involving: utilizing multi-agent technology, working collaboratively through a combination of different expert agents, to provide investors (i.e., users) with comprehensive and professional individual stock analysis recommendations.

[0086] Specifically, the solution provided in the embodiments of the present application can be as follows Figure 2 As shown, the following steps are included:

[0087] Step 21. Stock information collection:

[0088] Through various channels (such as financial platforms, news websites, social media, etc.), we comprehensively collect basic information of each stock (such as name, announcement, listed company, industry, shareholder ratio and other identity-related content), historical data (such as stock fluctuations - ups and downs), market dynamics (such as announcements, shareholders' meetings, news related to industry policies, research reports and news related to stocks, etc.) and other information texts, and store them in the database. These information texts can be texts from multiple sources and different structures (such as text, charts, time series data, etc.), which are not limited here.

[0089] Step 22. Obtain the target stock; corresponding to the identification information of the target object determined above:

[0090] The user inputs the target stock (such as the stock name and / or code); or inputs a sentence involving a certain stock, from which the name stock_name and the code stock_number of the target stock can be extracted.

[0091] Step 23. Data classification processing (such as obtaining data based on stock name, stock code, etc.):

[0092] 1) According to the name of the target stock stock_name, query the relevant information text (i.e., the text collected in step 21) from the database, filter out industry information (such as the industry to which the stock belongs, industry-related information (dynamics), etc.), text info_text related to stockholders' emotions (which may include the above-mentioned user feedback-related text), and further give the proportion of information related to optimism and pessimism according to the text info_text, i.e., the market sentiment score sentiment_score (corresponding to the above-mentioned market feedback information), and the interval of the score may be [-5,5]; this part of operation 1) may correspond to the above-mentioned query of information text from the database according to the object name to obtain the industry information and user feedback-related text; and obtain the market feedback information according to the user feedback-related text.

[0093] 2) Using the stock code stock_number of the target stock, the query service interface obtains multi-dimensional historical trends and technical indicators, etc., as the stock quantitative information stock_quantitative_information; corresponding to the above-mentioned query service interface according to the object index, the historical trend information and technical indicator information are obtained.

[0094] Step 24. Expert agent analysis:

[0095] Based on the big model, different expert agents are constructed using different prompts. The expert agents may include:

[0096] 1) Industry expert agent (corresponding to the first agent mentioned above): responsible for analyzing industry information, including industry policies, market dynamics, competitor conditions, etc., to evaluate the position and prospects (such as long-term trends) of the target stock in the industry; corresponding to the use of the first agent mentioned above, output information is obtained based on the industry information in the analysis reference information, and the output information includes: industry position information and prospect information corresponding to the target object;

[0097] 2) Psychological expert agent (corresponding to the second agent): combining the market sentiment score, evaluating the impact of market sentiment on the target stock, and predicting the future trend (such as the short-term trend); corresponding to the use of the second agent, based on the market feedback information in the analysis reference information, evaluating the impact of market feedback on the target object, and obtaining the first prediction information corresponding to the target object;

[0098] 3) Quantitative expert agent (corresponding to the third agent mentioned above): Based on the historical trend and technical indicators of the target stock (such as MACD, price-earnings ratio, etc.), use quantitative analysis methods to predict the long-term and short-term trends of its stock price; corresponding to the use of the third agent mentioned above, quantitative analysis is performed on the historical trend information and technical indicator information in the analysis reference information to obtain the second prediction information corresponding to the target object. Among them, "trend" can be, for example, ups and downs, but is not limited to this.

[0099] Among them, when constructing different expert agents (such as assigning roles, discourse weights, etc.), three experts can be set to speak in a random order, and each time the opinions of the previous experts, previous_expert_opinion, are spliced ​​into the prompt word of the next expert (equivalent to interaction between different experts). For example: set the field_target value to "chip", the stock_name value to "chip company", the info_text value to be omitted, the sentiment_score value to -3, and the initial value of previous_expert_opinion to be empty; if the random order is industry expert, psychological expert, and quantitative expert, the prompt word prompt can be constructed separately as follows:

[0100] (1) Industry expert prompt word prompt_1 = "Role: You are an industry expert in the field of {field_target}; Discourse weight: 0.2; Task: Please combine the relevant materials of {stock_name} stock to judge the future prospects or trends of the stock; Related information: {info_text}; The previous expert's opinion is {previous_expert_opinion}; Please give the stock prospects or trends: Industry experts believe that..."; The aforementioned field content can be called through the format format(field_target, stock_name, info_text, previous_expert_opinion), prompt_1 is input into the large model to obtain the output result result_1 (corresponding to the above output information), for example, result_1 = "Industry experts believe that: This stock belongs to the chip industry. Affected by the recent recovery of the industry, the subsequent trend of the stock is expected to rise"; Corresponding to the above-mentioned use of the first agent, according to the industry information in the analysis reference information, the output information is obtained, and the output information includes: the industry status information and prospect information corresponding to the target object;

[0101] (2) Psychologist prompt word prompt_2 = "Role: You are a psychologist in the field of {field_target}; Discourse weight: 0.3; Task: Please combine the relevant materials of {stock_name} stock to judge the future prospects or trends of the stock; Related information: {info_text}; Market sentiment score: positive: negative = {sentiment_score}; The previous expert's opinion is {previous_expert_opinion}; Please give the stock's prospects or trends: Psychologists believe that..."; The above field content can be called through format(field_target, stock_name, sentiment_score, previous_expert_opinion). In this case, previous_expert_opinion = result_1 (that is, the previous expert's opinion is used as the next expert's opinion). prompt); after inputting prompt_2 into the large model, the output result result_2 (corresponding to the above-mentioned first prediction information) is obtained, for example, result_2 = "Psychological experts believe that although industry experts believe that the recent recovery of the chip industry will affect the subsequent trend of the stock, the investor confidence in the stock is obviously insufficient. In the short term, market sentiment will suppress the stock trend. It is expected that the subsequent performance of the stock will show a downward trend"; wherein, the discourse weights of the first few experts can also be spliced ​​into the prompt word prompt of the next expert, so that the current expert can consider the output results of the first few in combination with the discourse weight, and then give its own output result; corresponding to the above-mentioned use of the second agent, according to the second information, and the market feedback information in the analysis reference information, the degree of influence of the market feedback on the target object is evaluated to obtain the first prediction information corresponding to the target object; the second information includes: the output information and the third weight corresponding to the output information;

[0102] (3) Quantitative expert prompt word prompt_3 = "Role: You are a quantitative expert in the field of {field_target}; Discourse weight: 0.5; Task: Please combine the relevant materials of {stock_name} stock to judge the future prospects or trends of the stock; Market sentiment score: positive: negative = {sentiment_score}; The opinions of the first two experts are {previous_expert_opinion}; Please give the stock outlook or trend: Quantitative experts believe that..."; The above field content can be called through format(field_target, stock_name, sentiment_score, previous_expert_opinion), in which case previous_expert_opinion = result_1 + result_2 (that is, the opinions of the first two experts are used as the input of the next expert prompt); prompt_3 After inputting the large model, the output result result_3 (corresponding to the above-mentioned second prediction information) is obtained, for example, result_3 = "Quantitative experts believe that the industry to which the stock belongs shows signs of recovery, and from a technical perspective, the stock price is in an upward channel. Although the current market sentiment tends to be bearish, the decline in stock prices (a 3% drop in the past two days) is only a short-term behavior. It is expected that the subsequent trend of the stock will continue to be bullish"; wherein, the discourse weights of the first few experts can also be spliced ​​into the prompt word prompt of the next expert, so that the current expert can consider the output results of the first few in combination with the discourse weight, and then give his own output result; corresponding to the above-mentioned use of the third agent, according to the third information, a quantitative analysis is performed on the historical trend information and technical indicator information in the analysis reference information to obtain the second prediction information corresponding to the target object; the second information includes: the output information and the third weight corresponding to the output information; and the first prediction information and the first weight corresponding to the first prediction information.

[0103] Step 25. The referee agent determines:

[0104] Based on the large model, a referee agent is constructed using prompts, and the output results of each expert agent are summarized to evaluate whether the opinions (corresponding to each output result) are consistent; corresponding to the use of the fourth agent, it is determined whether the output information, the first prediction information, and the second prediction information match; wherein the prompts of the referee agent are as follows:

[0105] prompt_judge = "Role: You are a referee expert; Task: Please judge whether the opinions of the three experts are consistent based on the opinions of the following experts in different fields; Industry expert opinion: {result_1}; Psychological expert opinion: {result_2}; Quantitative expert opinion: {result_3}; Please output the judgment result: "; You can call the above field content through format (result_1, result_2, result_3), input prompt_judge into the big model to get the output result result_judge, according to the result:

[0106] 1) If result_judge = "consistent", it means that the experts (agents) have the same opinion and can proceed to the next step (step 27) to summarize the suggestions; corresponding to the above-mentioned matching situation, according to the output information, the first prediction information, the second prediction information and the user feature information, the analysis result corresponding to the target object is obtained.

[0107] 2) If result_judge = "inconsistent", it means that there are inconsistencies in the opinions of the experts, and users can be invited to give their opinions.

[0108] It is explained here that whether the evaluation opinions are consistent can be performed in combination with the discourse weights of each expert agent. For example, if result_1 indicates a good prospect (discourse weight: 0.2), result_2 indicates a bad prospect (discourse weight: 0.3), and result_3 indicates a good prospect (discourse weight: 0.5), then the opinions can be evaluated to be consistent and the opinions indicate a good prospect, etc., but this is not limited to this.

[0109] Step 26. User opinion input:

[0110] The different opinions of the expert agents are displayed to the user, and the user is invited to express his or her user_opinion (for example, "I think stockholder sentiment is often a reverse indicator"), or to filter existing opinions and viewpoints, or to add more information about individual stocks; thereby obtaining the user's viewpoint (i.e., user opinion); corresponding to the above-mentioned situation where the output information, the first prediction information, and the second prediction information do not match, the user's opinion is obtained.

[0111] Step 27. Summarize the recommendations:

[0112] In combination with the opinions of each expert agent, user information (corresponding to the above user characteristic information, such as: the amount of funds user_funds, the maximum bearing capacity user_max_bearing, the individual stock preference user_preference, etc.), and (optional) user opinion (user_opinion), by constructing the prompt word prompt_summary_suggestion as the input of the big model, the big model is used to give an analysis summary (such as rise and fall) and investment advice (such as selling, holding or buying according to the characteristics of the stock, etc.) for the target individual stock. Corresponding to the above output information, the first prediction information, the second prediction information, the user opinion and the user characteristic information, the analysis result corresponding to the target object (which may include the analysis summary and investment advice) can be obtained; for example:

[0113] prompt_summary_suggestion="Task: Please summarize the opinions of experts in the following different fields and give investment suggestions based on user information; Industry expert opinions: {result_1}; Psychological expert opinions: {result_2}; Quantitative expert opinions: {result_3}; User opinion: {user_opinion}; User information: Funds {user_funds}, Maximum bearing capacity {user_max_bearing}, Stock preference {user_preference}; Please output the json result: {"Summary":; "Suggestion":}"; You can use the form at(result_1,result_2,result_3,user_opinion,user_funds,user_max_bearing,user_preference) calls the above fields. Here, user_funds = "holding 80,000, loss 20,000", user_max_bearing = "maximum retracement 50%", user_preference = "prefer to buy when the price rebounds, sell when the profit is 30%", and input prompt_summary_suggestion into the big model to get the output result (corresponding to the above analysis result), for example:

[0114] result_summary_suggestion = {

[0115] "Summary": "Combining the opinions of the three experts, industry experts and quantitative experts all believe that the stock is affected by the recent recovery of the chip industry and the subsequent trend is bullish. Although psychological experts believe that investors lack confidence and market sentiment may suppress the stock trend in the short term, from a technical perspective, the stock price is in an upward channel, and the recent decline is only a short-term behavior, and the decline is not large. Therefore, overall, the subsequent trend of the stock is still optimistic";

[0116] "Suggestion": "1. Considering that quantitative experts believe that the stock price is in an upward channel from a technical perspective, you can buy when the stock price falls back in the short term. 2. The recent stock price decline is only 3%, so it is recommended to buy on dips."

[0117] }.

[0118] It is explained here that the summary suggestions can be implemented in combination with the discourse weights of each expert agent. For example, if result_1 indicates a good prospect (discourse weight: 0.2), result_2 indicates a bad prospect (discourse weight: 0.3), and result_3 indicates a good prospect (discourse weight: 0.5), then the conclusion can be drawn: the trend is bullish, but this is not limited to this.

[0119] For the above method, the embodiment of the present application can be used Figure 3 The functional framework shown in the figure is implemented, which specifically involves the following modules:

[0120] 1. The information crawling module can be used to: comprehensively collect basic information, historical data, market trends, etc. of each stock through various channels and store them in the database.

[0121] 2. The information classification module can be used to: query relevant information text from the database based on the name of the target stock, filter out text related to industry information and stockholder sentiment, and give a market sentiment score.

[0122] 3. The stock price query module can be used to: use the stock code of the target stock to query the service interface to obtain multi-dimensional historical trends and technical indicators, etc.

[0123] 4. Expert analysis module can be used to: use the big model to build a variety of different agents, analyze from the professional perspectives of industry experts, psychological experts, quantitative experts, etc., and give their respective expert opinions.

[0124] 5. The referee judgment module can be used to: summarize the output results of each expert agent, evaluate whether the opinions are consistent, and decide whether to invite users to give their opinions.

[0125] 6. Summary and suggestion module can be used to: combine the opinions of expert agents (and user opinions) to give analysis summaries and investment suggestions for target stocks.

[0126] From the above, the solution provided by the embodiment of the present application mainly includes:

[0127] 1) Processing and analysis of multi-source heterogeneous data and multi-dimensional data: Use industry information, text data related to stockholder sentiment, and stock data to make relevant judgments. In addition, more dimensions can be added to show users more comprehensive information. This is not limited here;

[0128] 2) Multiple expert agent analysis: Based on the big model and prompts, multiple agent roles such as industry experts, psychological experts and quantitative experts are constructed, and their discourse weights are given. Combined with the expert interaction method, professional analysis of target stocks is carried out from the aspects of industry information analysis, market sentiment influence and stock price trend;

[0129] 3) Judgement of the referee agent and interaction with the user: The referee agent can decide whether to invite the user to give his or her opinion based on the consistency of opinions of the expert agents, thus realizing human-computer interaction.

[0130] In summary, the solution provided in the embodiments of the present application can have the following beneficial effects:

[0131] 1. Comprehensiveness of information integration and analysis:

[0132] Through multi-agent technology, we integrate multi-dimensional information such as stock price trends, industry information, market sentiment, etc. to provide more comprehensive stock analysis reports.

[0133] 2. Improved professionalism and synergy of expert systems:

[0134] This scheme uses different agents to process different types of information respectively, and the referee agent conducts summary analysis, which improves the relevance between information and the accuracy of comprehensive analysis.

[0135] 3. Improvement of user engagement:

[0136] When expert opinions are inconsistent, user perspectives are introduced to make the analysis results more in line with user needs and improve user participation and satisfaction.

[0137] It is explained here that this solution can be applied to the field of natural language processing (NLP) to achieve collaborative work of multi-agent technology based on a large model, but it is not limited to this.

[0138] The present application also provides an information processing device, such as Figure 4 As shown, including:

[0139] A first determination module 41, used to determine identification information of a target object;

[0140] A first acquisition module 42 is used to acquire analysis reference information corresponding to the target object according to the identification information; the analysis reference information includes: industry information, market feedback information, historical trend information, and technical indicator information;

[0141] The second determination module 43 is used to determine the analysis result corresponding to the target object according to the analysis reference information.

[0142] The information processing device provided in the embodiment of the present application determines the identification information of the target object; based on the identification information, obtains the analysis reference information corresponding to the target object; the analysis reference information includes: industry information, market feedback information, historical trend information and technical indicator information; based on the analysis reference information, determines the analysis result corresponding to the target object; can support the analysis based on more comprehensive information, and comprehensively consider the relevant information to ensure the correlation between the information in order to form a comprehensive judgment, thereby providing comprehensive and professional analysis results, improving the low analysis accuracy, and well solving the problem of low analysis accuracy in related solutions for data analysis in the prior art.

[0143] Among them, determining the analysis result corresponding to the target object based on the analysis reference information includes: using a first agent to obtain output information based on the industry information in the analysis reference information, and the output information includes: industry status information and prospect information corresponding to the target object; using a second agent to evaluate the impact of market feedback on the target object based on the market feedback information in the analysis reference information, and obtain first prediction information corresponding to the target object; using a third agent to perform quantitative analysis on the historical trend information and technical indicator information in the analysis reference information, and obtain second prediction information corresponding to the target object; obtaining the analysis result corresponding to the target object based on the output information, the first prediction information and the second prediction information.

[0144] In an embodiment of the present application, (1) the use of the first agent to obtain output information based on the industry information in the analysis reference information includes: using the first agent to obtain output information based on the first information and the industry information in the analysis reference information; the first information includes: the first prediction information and the first weight corresponding to the first prediction information; and / or, the second prediction information and the second weight corresponding to the second prediction information; and / or, (2) the use of the second agent to evaluate the impact of market feedback on the target object based on the market feedback information in the analysis reference information to obtain the first prediction information corresponding to the target object includes: using the second agent to evaluate the impact of market feedback on the target object based on the second information and the market feedback information in the analysis reference information degree, to obtain the first prediction information corresponding to the target object; the second information includes: the output information and the third weight corresponding to the output information; and / or, the second prediction information and the second weight corresponding to the second prediction information; and / or, (3) the use of a third agent to perform a quantitative analysis on the historical trend information and technical indicator information in the analysis reference information to obtain the second prediction information corresponding to the target object, including: using a third agent, based on the third information, to perform a quantitative analysis on the historical trend information and technical indicator information in the analysis reference information to obtain the second prediction information corresponding to the target object; the second information includes: the output information and the third weight corresponding to the output information; and / or, the first prediction information and the first weight corresponding to the first prediction information.

[0145] Among them, obtaining the analysis result corresponding to the target object based on the output information, the first prediction information and the second prediction information includes: using a fourth agent to determine whether the output information, the first prediction information and the second prediction information match; if they match, obtaining the analysis result corresponding to the target object based on the output information, the first prediction information and the second prediction information and user feature information.

[0146] Furthermore, the information processing device also includes: a second acquisition module, used to obtain user opinions when the output information, the first prediction information and the second prediction information do not match; obtaining the analysis results corresponding to the target object based on the output information, the first prediction information and the second prediction information includes: obtaining the analysis results corresponding to the target object based on the output information, the first prediction information, the second prediction information, user opinions and user feature information.

[0147] In an embodiment of the present application, the identification information includes: an object name and an object index; obtaining analysis reference information corresponding to the target object based on the identification information includes: querying information text from a database based on the object name to obtain the industry information and user feedback related text; obtaining the market feedback information based on the user feedback related text; querying a service interface based on the object index to obtain the historical trend information and technical indicator information.

[0148] Among them, the implementation embodiments of the above-mentioned information processing method are all applicable to the embodiments of the information processing device and can achieve the same technical effects.

[0149] An embodiment of the present application also provides an information processing device, including a memory, a processor, and a program stored in the memory and executable on the processor; when the processor executes the program, the above-mentioned information processing method is implemented.

[0150] Among them, the implementation embodiments of the above-mentioned information processing method are all applicable to the embodiments of the information processing device and can achieve the same technical effects.

[0151] An embodiment of the present application also provides a readable storage medium on which a program is stored, and when the program is executed by a processor, the steps in the above-mentioned information processing method are implemented.

[0152] Among them, the implementation embodiments of the above-mentioned information processing method are all applicable to the embodiments of the readable storage medium and can achieve the same technical effects.

[0153] The embodiment of the present application also provides a computer program product, including computer instructions. When the computer instructions are executed by a processor, the various processes of the method embodiment of the above-mentioned information processing method are implemented, and the same technical effect can be achieved. To avoid repetition, they are not repeated here.

[0154] It should be noted that many functional components described in this specification are referred to as modules in order to more particularly emphasize the independence of their implementation methods.

[0155] In the embodiment of the present application, module can be implemented with software so that it can be executed by various types of processors. For example, an executable code module of an identification can include one or more physical or logical blocks of computer instructions, for example, it can be constructed as an object, process or function. Nevertheless, the executable code of the identified module does not need to be physically located together, but can include different instructions stored in different positions, and when these instructions are logically combined together, it constitutes a module and realizes the specified purpose of the module.

[0156] In fact, executable code module can be a single instruction or many instructions, and can even be distributed on a plurality of different code segments, distributed among different programs, and distributed across a plurality of memory devices. Similarly, operating data can be identified in the module, and can be implemented and organized in the data structure of any appropriate type according to any appropriate form. The operating data can be collected as a single data set, or can be distributed in different locations (including on different storage devices), and can only be present on a system or network as an electronic signal at least in part.

[0157] When a module can be implemented by software, considering the level of existing hardware technology, a person skilled in the art can build a corresponding hardware circuit to implement the corresponding function of the module that can be implemented by software without considering the cost. The hardware circuit includes a conventional very large scale integration (VLSI) circuit or gate array and existing semiconductors such as logic chips, transistors, or other discrete components. The module can also be implemented by a programmable hardware device, such as a field programmable gate array, a programmable array logic, a programmable logic device, etc.

[0158] The above is a preferred embodiment of the present application. It should be pointed out that for ordinary personnel in this technical field, several improvements and modifications can be made without departing from the principles described in the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. An information processing method, characterized in that: include: Determine identification information of the target object; According to the identification information, obtaining analysis reference information corresponding to the target object; The analysis reference information includes: industry information, market feedback information, historical trend information and technical indicator information; An analysis result corresponding to the target object is determined according to the analysis reference information.

2. The information processing method according to claim 1, characterized in that: Determining the analysis result corresponding to the target object according to the analysis reference information includes: Using the first agent, according to the industry information in the analysis reference information, output information is obtained, wherein the output information includes: industry status information and prospect information corresponding to the target object; Using a second agent, based on the market feedback information in the analysis reference information, assess the degree of influence of the market feedback on the target object, and obtain first prediction information corresponding to the target object; Using a third agent, quantitatively analyzing the historical trend information and the technical indicator information in the analysis reference information to obtain second prediction information corresponding to the target object; An analysis result corresponding to the target object is obtained according to the output information, the first prediction information and the second prediction information.

3. The information processing method according to claim 2, characterized in that: The first agent is used to obtain output information according to the industry information in the analysis reference information, including: Using a first agent, output information is obtained according to first information and industry information in the analysis reference information; the first information includes: the first prediction information and a first weight corresponding to the first prediction information; and / or the second prediction information and a second weight corresponding to the second prediction information; And / or, the using of the second agent to evaluate the degree of influence of market feedback on the target object according to the market feedback information in the analysis reference information to obtain the first prediction information corresponding to the target object includes: Using the second agent, based on the second information and the market feedback information in the analysis reference information, the degree of influence of the market feedback on the target object is evaluated to obtain the first prediction information corresponding to the target object; the second information includes: the output information and the third weight corresponding to the output information; and / or, the second prediction information and the second weight corresponding to the second prediction information; And / or, the using of a third agent to perform quantitative analysis on the historical trend information and technical indicator information in the analysis reference information to obtain second prediction information corresponding to the target object includes: Using a third agent, based on the third information, a quantitative analysis is performed on the historical trend information and technical indicator information in the analysis reference information to obtain second prediction information corresponding to the target object; the second information includes: the output information and the third weight corresponding to the output information; and / or, the first prediction information and the first weight corresponding to the first prediction information.

4. The information processing method according to claim 2, characterized in that: Obtaining an analysis result corresponding to the target object according to the output information, the first prediction information, and the second prediction information includes: Determining, using a fourth agent, whether the output information, the first prediction information, and the second prediction information match; In the case of a match, an analysis result corresponding to the target object is obtained according to the output information, the first prediction information, the second prediction information and the user characteristic information.

5. The information processing method according to claim 4, characterized in that: Also includes: In the case where the output information, the first prediction information, and the second prediction information do not match, obtaining a user opinion; Obtaining an analysis result corresponding to the target object according to the output information, the first prediction information, and the second prediction information includes: An analysis result corresponding to the target object is obtained according to the output information, the first prediction information, the second prediction information, the user opinion and the user characteristic information.

6. The information processing method according to claim 1, characterized in that: The identification information includes: object name and object index; The acquiring, according to the identification information, analysis reference information corresponding to the target object includes: According to the object name, query the information text from the database to obtain the industry information and user feedback related text; Obtaining the market feedback information according to the user feedback related text; According to the object index, the service interface is queried to obtain the historical trend information and technical indicator information.

7. An information processing device, characterized in that: include: A first determination module, used to determine identification information of a target object; A first acquisition module, configured to acquire analysis reference information corresponding to the target object according to the identification information; The analysis reference information includes: industry information, market feedback information, historical trend information and technical indicator information; The second determination module is used to determine the analysis result corresponding to the target object according to the analysis reference information.

8. An information processing device, comprising a memory, a processor, and a program stored in the memory and executable on the processor; characterized in that: When the processor executes the program, the information processing method according to any one of claims 1 to 6 is implemented.

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

10. A computer program product, characterized in that The method comprises computer instructions, which, when executed by a processor, implement the steps of the information processing method according to any one of claims 1 to 6.

Images