Multilingual fine-grained emotion detection for business intelligence

Start - End 
2017 - 2022 (ongoing)
Type 
Department(s) 
Department of Translation, Interpreting and Communication

Tabgroup

Abstract

Despite the plethora of often visually very attractive social media monitoring tools currently available on the market, none of the existing solutions provide fine-grained insights to support reputation monitoring. The salient dimensions of an organization’s reputation may differ across stakeholders: employees, investors, or consumers, who are all interested in different corporate performance characteristics (e.g. financial performance for investors, workplace condition for potential employees, quality of products and services for customers, social responsibility for a wider public). To have a better understanding of how stakeholders in different countries respond to companies and other organizations, we investigate different emotion frameworks to model human emotions expressed online to support business intelligence. A learnable emotion framework will be operationalized and implemented in an automatic emotion detection system for reputation monitoring.