Combat Greenwashing with GoalSpotter: Automatic Sustainability Objective Detection in Heterogeneous

Link: https://dl.acm.org/doi/10.1145/3627673.3680110

Conference: CIKM 2024 Applied Research Papers

Keywords: Greenwashing, Heterogeneous Reports, Text Mining, Transforer

Summary

Build a clustering-based algorithm to extract text elements of heterogeneous sustainability reports (e.g., pdf, html) and group them into informative text blocks. Then build a transforer-based model to detect sustainability objectives in the text blocks.

alt text

Methodology

  1. Discrete textual content processing Solution: clustering

  2. Label

  3. Fine tune Transforer

PreviousPredicting Impression Evaluation of Building Exterior Appearance Using Street Image Big Data and DeeNextA Multi-task Model for Sentiment Aided Stance Detection of Climate Change Tweets

Last updated