SkimLit

Overview

A lot of users face the problem of going through lengthy research papers just to get some little information that they are looking for. Even with abstracts summarizing the contents of the paper in a paragraph, it is hard to understand the context sometime. SkimLit takes in the abstract of any research paper and breaks it down into labelled segments. These labels can be any one of the following five - objectives, background, methods, results, and conclusions. With these labels preceding the sentence of the abstract, it becomes a breeze to read through the content which would otherwise take longer to extract from.

Technologies Used

Key Learnings

• Advanced NLP techniques
• Encoding processes - with Universal Sentence Encoder (USE) & GloVe Embeddings
• Multi-modality in neural networks