Cisco at AAAI-CAD21 shared task: Predicting Emphasis in Presentation Slides using Contextualized Embeddings
This paper addresses the problem of predicting word emphasis in presentation slides, which could be useful for automated presentation generation or accessibility tools.
The authors tackled the problem of predicting emphasis for each word in a presentation slide. They achieved a score of 0.518 on the evaluation leaderboard, ranking 3rd, and 0.543 on the post-evaluation leaderboard, ranking 1st.
This paper describes our proposed system for the AAAI-CAD21 shared task: Predicting Emphasis in Presentation Slides. In this specific task, given the contents of a slide we are asked to predict the degree of emphasis to be laid on each word in the slide. We propose 2 approaches to this problem including a BiLSTM-ELMo approach and a transformers based approach based on RoBERTa and XLNet architectures. We achieve a score of 0.518 on the evaluation leaderboard which ranks us 3rd and 0.543 on the post-evaluation leaderboard which ranks us 1st at the time of writing the paper.