Unimodal Intermediate Training for Multimodal Meme Sentiment Classification
This addresses the challenge of limited labelled data for multimodal meme sentiment classification, which is an incremental improvement for researchers and developers in social media analysis.
The paper tackled the problem of sentiment classification for multimodal memes by supplementing training with unimodal data, resulting in a statistically significant performance improvement and a 40% reduction in required labelled memes without performance loss.
Internet Memes remain a challenging form of user-generated content for automated sentiment classification. The availability of labelled memes is a barrier to developing sentiment classifiers of multimodal memes. To address the shortage of labelled memes, we propose to supplement the training of a multimodal meme classifier with unimodal (image-only and text-only) data. In this work, we present a novel variant of supervised intermediate training that uses relatively abundant sentiment-labelled unimodal data. Our results show a statistically significant performance improvement from the incorporation of unimodal text data. Furthermore, we show that the training set of labelled memes can be reduced by 40% without reducing the performance of the downstream model.