Automated Driving Force Identification from Financial Text
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional scenario creation for corporate strategy planning requires expertise and is time-consuming, making it difficult for non-specialists to identify driving forces relevant to the company's situation.
Innovation Solution
A strategy planning support device that performs EPU analysis using frequently occurring words from a company's financial statement, calculates uncertainty using the Dirichlet allocation method, detects relevant words using cosine similarity, and determines high-order specific words as elements of driving forces.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual expert analysis is used to determine driving forces, then expertise and experience are leveraged, but the process becomes time-consuming and difficult for non-specialists
Solution Approach 1:
The patent introduces an automated analysis system that acts as an intermediary between raw financial data and driving force identification. This system uses natural language processing and machine learning algorithms to automatically extract and analyze driving forces from financial statements, eliminating the need for manual expert analysis while maintaining accuracy.
Solution Approach 2:
The patent replaces the manual mechanical process of expert analysis with an automated computational system. The system uses algorithms to process financial data, identify patterns, and determine driving forces automatically, substituting human expert time with machine-based analysis that is both fast and reproducible.
2Reliability
If manual expert workshops are conducted to identify driving forces, then deep domain knowledge is applied, but the complexity and difficulty for non-specialists increases
Solution Approach 1:
The patent enables the financial analysis system to perform self-service by automatically processing financial statements and identifying driving forces without requiring external expert intervention. The system uses pre-trained models and algorithms to autonomously complete the analysis task, reducing the complexity of the overall process while maintaining reliable results.
Solution Approach 2:
The patent transforms the complex qualitative expert judgment process into quantifiable parameters that can be automatically processed. By converting driving force identification into measurable metrics extracted from financial data, the system reduces process complexity while maintaining the reliability needed for quality scenario planning.
3Adaptability or versatility
If conventional scenario planning methods are used, then general framework is available, but customization to specific company situations becomes difficult
Solution Approach 1:
The patent segments the scenario planning process into distinct components: data collection from financial statements, driving force identification, uncertainty analysis, and scenario generation. This segmentation allows the system to be customized for different companies by simply changing the input data while maintaining the robust general framework, thereby improving both adaptability and company-specific relevance.
Data Source
AI summary
An object of the present disclosure is to propose a technology that can allow even non-specialists to easily identify driving forces (DF) in order to support the creation of scenarios in line with the actual situation of a company.A strategy planning support device of the present disclosure includes: an EPU analysis unit that performs EPU analysis on a word in a social situation survey document by using a frequently occurring word in a financial statement of an organization as a Policy classification term; an uncertainty calculation unit that calculates uncertainty of each classification analyzed by the EPU analysis unit by using a potential Dirichlet allocation method; a relevance calculation unit that detects a predetermined word appearing around the word analyzed by the EPU analysis unit in a specific classification having the highest uncertainty as calculated by the uncertainty calculation unit, and calculates relevance of the predetermined word using a cosine similarity; a determination unit that determines a high-order specific word having a high relevance value calculated by the relevance calculation unit, as an element of each coordinate axis of a driving force; and an output unit that outputs the specific word determined by the determination unit.