CN-122022903-A - Industrial investment automatic analysis method and device

CN122022903ACN 122022903 ACN122022903 ACN 122022903ACN-122022903-A

Abstract

The application discloses an industrial investment automatic analysis method and device, and relates to the technical field of finance. The method comprises the steps of obtaining a target object to be analyzed and an analysis report template corresponding to the target object to be analyzed, splitting a target analysis task corresponding to the target object to be analyzed into a plurality of subtasks according to the analysis report template by adopting a preset framework, determining an analysis result corresponding to each subtask based on a task analysis mode corresponding to each subtask, and carrying out templated integration on the plurality of analysis results according to the analysis report template to obtain a target analysis report. The method solves the problems that the traditional method for automatically analyzing the industrial investment based on manual work has low information analysis and integration efficiency, weak association capability and strong subjectivity, and is easy to influence the rationality of investment decisions, improves the efficiency of information analysis and integration, improves the association capability, and avoids the problem that the rationality of investment decisions is possibly influenced by the subjective influence of the manual work.

Inventors

SHI YANYING
WANG HUAN
HUANG WEN

Assignees

紫金诚征信有限公司

Dates

Publication Date: 20260512
Application Date: 20251230

Claims (10)

1. An industrial investment automation analysis method, comprising: acquiring a target object to be analyzed and an analysis report template corresponding to the target object to be analyzed; Splitting a target analysis task corresponding to the target object to be analyzed into a plurality of subtasks by adopting a preset framework according to the analysis report template; Determining an analysis result corresponding to each subtask based on a task analysis mode corresponding to each subtask; and carrying out template integration on a plurality of analysis results according to the analysis report template to obtain a target analysis report.
2. The method of claim 1, wherein determining an analysis result corresponding to each subtask based on a task analysis manner corresponding to each subtask comprises: Generating a first analysis text corresponding to the target object to be analyzed based on a first text generation subtask; generating subtasks based on the second index files, determining an industrial data knowledge base corresponding to the target object to be analyzed, and determining a plurality of index files based on the industrial data knowledge base; generating an index description text corresponding to each index file according to a plurality of index files based on a third text generation subtask, inputting the index files and the index description texts into an index analysis model, and outputting a second analysis text corresponding to the target object to be analyzed; And generating a plurality of visual charts according to a plurality of index files based on a fourth chart generation subtask.
3. The method of claim 2, wherein templated integration of the plurality of analysis results according to the analysis report template to obtain a target analysis report comprises templated integration of the plurality of visualization charts, the first analysis text, and the second analysis text to obtain a target analysis report.
4. The method of claim 2, wherein generating a first analysis text corresponding to the target object to be analyzed based on a first text generation subtask comprises: Extracting target industries to be analyzed in the target objects to be analyzed, and carrying out text supplementation on the target industries to be analyzed by adopting a preset text supplementation mode to obtain target texts to be analyzed; determining a plurality of document block types according to the target text to be analyzed, and determining a plurality of document blocks from an industrial data knowledge base according to the target text to be analyzed and the plurality of document block types; respectively splicing a plurality of document block types with the corresponding document blocks to obtain a plurality of spliced document blocks, and carrying out text fusion on the spliced document blocks and the target text to be analyzed to obtain a fusion text; and inputting the fusion text into a fusion analysis model, and outputting the first analysis text.
5. The method of claim 4, wherein inputting the fused text into an analytical model and outputting the first analytical text comprises: Acquiring main document metadata information corresponding to each document block in the fusion text; Determining the weight of the document block corresponding to each document block according to the metadata information of the plurality of main documents; and carrying out fusion analysis on the fusion text according to the document block weight corresponding to each document block to obtain the first analysis text.
6. The method of claim 5, wherein determining a document block weight for each of the document blocks based on a plurality of the master document metadata information comprises: Determining a document authority index of a main document corresponding to each document block according to the main document metadata information; determining a document integrity index of a main document corresponding to each document block according to the main document metadata information, wherein the document integrity index is used for indicating the number of entities contained in the main document; acquiring authority superparameter corresponding to the document authority index and integrity superparameter corresponding to the document integrity index; and determining a plurality of document block weights according to a plurality of authority indexes, a plurality of integrity indexes, the authority super-parameters and the integrity super-parameters.
7. The method of claim 2, wherein determining an industry data knowledge base corresponding to the target object to be analyzed and determining a plurality of index files based on the industry data knowledge base comprises: extracting a target enterprise to be analyzed in the target object to be analyzed, and determining an industry data knowledge base corresponding to the target enterprise to be analyzed; acquiring enterprise research and development data and enterprise patent data from the industrial data knowledge base, and calculating scientific research capability indexes based on the enterprise research and development data and the enterprise patent data to obtain scientific research capability index files; and acquiring asset data from the industrial data knowledge base, and calculating key financial ratio indexes based on the asset data to obtain a financial ratio index file.
8. The method of claim 1, further comprising, prior to obtaining a target object to be analyzed and an analysis report template corresponding to the target object to be analyzed: Acquiring a plurality of main documents from a plurality of data sources, and acquiring document content and document metadata information corresponding to each main document; converting a plurality of main documents into structured data with a preset list structure according to the document content and the document metadata information corresponding to each main document; respectively carrying out document blocking and metadata inheritance on a plurality of main documents according to the structured data corresponding to each main document to obtain a plurality of document sub-blocks; And storing a plurality of the document sub-blocks into an industry data knowledge base.
9. The method of claim 8, wherein performing document blocking and metadata inheritance on the plurality of main documents according to the structured data corresponding to each main document to obtain a plurality of document sub-blocks, respectively, includes: Determining a plurality of document block types corresponding to each main document according to the document content in each structured data; Respectively carrying out document block cutting on a plurality of main documents according to the document block type corresponding to each main document to obtain a plurality of reference document sub-blocks; and respectively inheriting the document metadata information in each structured data to the reference document sub-blocks corresponding to the structured data to obtain a plurality of document sub-blocks.
10. An industrial investment automation analysis device, comprising: the acquisition unit is used for acquiring a target object to be analyzed and an analysis report template corresponding to the target object to be analyzed; The task decomposition unit is used for decomposing a target analysis task corresponding to the target object to be analyzed into a plurality of subtasks according to the analysis report template by adopting a preset framework; the determining unit is used for determining an analysis result corresponding to each subtask based on a task analysis mode corresponding to each subtask; and the integrating unit is used for carrying out templated integration on the analysis results according to the analysis report template to obtain a target analysis report.

Description

Industrial investment automatic analysis method and device Technical Field The application relates to the technical field of finance, in particular to an industrial investment automatic analysis method and device. Background In the field of investment project evaluation, in the current domestic equity investment pattern, the traditional marketized investment institutions and state-run assets background institutions respectively bear the roles of GP and LP, in recent years, along with the increasingly prominent leading role of state-run assets in industrial investment, the traditional LP limiting positioning has hardly met the investment requirement of domestic middle-high-end industrial transformation upgrading, state-run assets LP is gradually transformed to a mixed mode of 'LP+GP', and the industrial investment capacity of state-run assets background institutions is more required. The choice of quality targets at the right moment is a manifestation of the industrial investment ability, whereas quality projects are scarce in the seller market, such as hard technology and high-tech technologies. This requires that the analysis report must be output quickly and structurally. Traditional investment analysis is highly dependent on personal experience of a foundation manager, and extraction, association and integration of various information (including production reports, enterprise annual reports, business plans and due-job investigation materials) are mainly finished manually. This leads to analytical conclusions, especially the work of the primary industry manager, with the general problem of "low information integration efficiency, weak correlation ability, and strong subjectivity". Because of personnel experience differences, analytical evaluations tend to be highly subjective and may impact the rationality of investment decisions. In the current large-scale investment management era, the mode of manually collecting and analyzing external information is difficult to meet the requirement of state-run assets investment institutions on time-consuming and analysis quality-heavy pre-investment management. Disclosure of Invention The application mainly aims to provide an industrial investment automatic analysis method and device, which are used for solving the problems that the traditional manual-based industrial investment automatic analysis mode is low in information analysis and integration efficiency, weak in association capability and strong in subjectivity, and the rationality of investment decisions is easily influenced, improving the efficiency of information analysis and integration, improving the association capability, and avoiding the problem that the rationality of investment decisions is possibly influenced by the subjective influence of people. In order to achieve the above objective, according to one aspect of the embodiments of the present application, an industrial investment automation analysis method is provided, which includes obtaining a target object to be analyzed and an analysis report template corresponding to the target object to be analyzed, splitting a target analysis task corresponding to the target object to be analyzed into a plurality of subtasks according to the analysis report template by using a preset framework, determining an analysis result corresponding to each subtask based on a task analysis mode corresponding to each subtask, and performing template integration on the plurality of analysis results according to the analysis report template to obtain a target analysis report. According to another aspect of the embodiment of the application, an industrial investment automation analysis device is provided, which comprises an acquisition unit, a task decomposition unit, a determination unit and an integration unit, wherein the acquisition unit is used for acquiring a target object to be analyzed and an analysis report template corresponding to the target object to be analyzed, the task decomposition unit is used for dividing a target analysis task corresponding to the target object to be analyzed into a plurality of subtasks according to the analysis report template by adopting a preset framework, the determination unit is used for determining an analysis result corresponding to each subtask based on a task analysis mode corresponding to each subtask, and the integration unit is used for carrying out template integration on the plurality of analysis results according to the analysis report template to obtain a target analysis report. According to yet another aspect of an embodiment of the present application, there is provided a computer-readable storage medium storing computer instructions for causing a computer to perform the industrial investment automation analysis method as above. According to yet another aspect of an embodiment of the present application, there is provided an electronic device including at least one Central Processing Unit (CPU) for system scheduling,