AdCOFE: Advanced Contextual Feature Extraction in conversations for emotion classification

Vaibhav Bhat; Anita Yadav; Sonal Yadav; Dhivya Chandrasekaran; Vijay Mago

doi:10.7717/peerj-cs.786

AdCOFE: Advanced Contextual Feature Extraction in conversations for emotion classification

PeerJ Comput Sci. 2021 Dec 9:7:e786. doi: 10.7717/peerj-cs.786. eCollection 2021.

Authors

Vaibhav Bhat¹, Anita Yadav¹, Sonal Yadav¹, Dhivya Chandrasekaran¹, Vijay Mago¹

Affiliation

¹ Department of Computer Science, Lakehead University, Thunderbay, Ontario, Canada.

Abstract

Emotion recognition in conversations is an important step in various virtual chatbots which require opinion-based feedback, like in social media threads, online support, and many more applications. Current emotion recognition in conversations models face issues like: (a) loss of contextual information in between two dialogues of a conversation, (b) failure to give appropriate importance to significant tokens in each utterance, (c) inability to pass on the emotional information from previous utterances. The proposed model of Advanced Contextual Feature Extraction (AdCOFE) addresses these issues by performing unique feature extraction using knowledge graphs, sentiment lexicons and phrases of natural language at all levels (word and position embedding) of the utterances. Experiments on emotion recognition in conversations datasets show that AdCOFE is beneficial in capturing emotions in conversations.

Keywords: Chatbots; Dyadic conversation; Emotion recognition.

Grants and funding

Publication costs are funded by a NSERC Discovery Grant (RGPIN-2017-05377), held by Dr. Vijay Mago. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.