ESG Supply Chain Forecasting via Sentiment Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a lack of a unified method to effectively forecast raw material demand considering Environmental, Social, and Governance (ESG) factors, which are crucial for investors making informed decisions and aligning with sustainable development goals, as existing methods do not adequately account for the holistic impact of ESG criteria on supply and demand.
Innovation Solution
A computer-implemented method and system that retrieves news articles and industry reports related to raw material suppliers and specified materials, generates sentiment scores using a sentiment scoring model, and combines these scores with supply chain and ESG data to forecast demand through an ESG forecast model, integrating ESG criteria into the forecasting process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional forecasting methods are used, then the forecasting process is simple, but ESG factors are not adequately considered
Solution Approach 1:
The system segments the forecasting process into distinct modules: news article retrieval, vectorization, sentiment scoring, and ESG forecast modeling. Each module handles a specific aspect of the forecasting task, allowing the complex ESG analysis to be broken down into manageable components that can be developed and maintained independently
Solution Approach 2:
The patent introduces intermediary components including a vectorization layer that converts news articles into numerical representations, and a sentiment scoring model that bridges raw text data and ESG forecast outcomes. These intermediaries enable the integration of unstructured news data with structured forecasting models
2Loss of information
If ESG factors are comprehensively considered, then investment decisions are more informed, but the forecasting process becomes more complex
Solution Approach 1:
The system performs preliminary actions by retrieving and vectorizing news articles before the main forecasting process. The sentiment scoring model pre-processes the vectorized data to extract sentiment information, preparing the data in advance for the ESG forecast model. This staged approach ensures comprehensive information capture while organizing the complexity into sequential steps
Solution Approach 2:
The patent transforms unstructured news article text into a different dimension (numerical vectors) through vectorization, then further transforms these vectors into sentiment scores. This dimensional transformation allows the system to process comprehensive text information using mathematical operations suitable for machine learning models
Data Source
AI summary
A method of forecasting raw material demand is provided. The method comprises retrieving news articles related to raw material suppliers as well as raw material industry articles related to specified raw materials. The news articles are classified as either ESG articles or non-ESG articles. The news articles and raw material industry articles are vectorized, and a subset of relevant articles are selected from the non-ESG articles. The ESG articles, relevant non-ESG articles, and raw material industry articles are fed into a sentiment scoring model, which generates sentiment scores for the raw material suppliers as well as sentiment scores for the raw materials. The sentiment scores are fed into an ESG forecast model along with supply chain data related to the raw materials and ESG data related to the raw material suppliers. The ESG forecast model then forecasts demand for the raw material suppliers to supply the raw materials.