A Multi-modal Approach to Fine-grained Opinion Mining on Video Reviews

TL;DR

This paper introduces a multi-modal method for extracting detailed opinions from video reviews by analyzing audio, video, and transcriptions, improving understanding over text-only approaches.

Contribution

It presents a novel multi-modal framework that operates at the sentence level without time annotations, integrating audio, video, and language features for fine-grained opinion mining.

Findings

Multi-modal approach outperforms text-only baselines.

Leveraging audio and video improves opinion detection accuracy.

Method effective on two distinct datasets.

Abstract

Despite the recent advances in opinion mining for written reviews, few works have tackled the problem on other sources of reviews. In light of this issue, we propose a multi-modal approach for mining fine-grained opinions from video reviews that is able to determine the aspects of the item under review that are being discussed and the sentiment orientation towards them. Our approach works at the sentence level without the need for time annotations and uses features derived from the audio, video and language transcriptions of its contents. We evaluate our approach on two datasets and show that leveraging the video and audio modalities consistently provides increased performance over text-only baselines, providing evidence these extra modalities are key in better understanding video reviews.