Entity-level Sentiment Analysis in Contact Center Telephone Conversations

TL;DR

This paper presents two approaches for entity-level sentiment analysis in contact center telephone conversations, using transformer-based and neural network models to extract business insights from transcripts.

Contribution

It introduces a system for entity-level sentiment analysis in English contact center transcripts, comparing transformer and CNN-based methods.

Findings

Transformer-based approach achieves higher accuracy.

CNN with heuristics offers a lightweight alternative.

System provides valuable business insights.

Abstract

Entity-level sentiment analysis predicts the sentiment about entities mentioned in a given text. It is very useful in a business context to understand user emotions towards certain entities, such as products or companies. In this paper, we demonstrate how we developed an entity-level sentiment analysis system that analyzes English telephone conversation transcripts in contact centers to provide business insight. We present two approaches, one entirely based on the transformer-based DistilBERT model, and another that uses a convolutional neural network supplemented with some heuristic rules.