PSU at CLEF-2020 ARQMath Track: Unsupervised Re-ranking using Pretraining

This paper elaborates on our submission to the ARQMath track at CLEF 2020. Our primary run for the main Task-1: Question Answering uses a two-stage retrieval technique in which the first stage is a fusion of traditional BM25 scoring and tf-idf with cosine similarity-based retrieval while the second stage is a finer re-ranking technique using contextualized embeddings. For the re-ranking we use a pre-trained roberta-base model (110 million parameters) to make the language model more math-aware. Our approach achieves a higher NDCG⁰ score than the baseline, while our MAP and P@10 scores are competitive, performing better than the best submission (MathDowsers) for text and text+formula dependent topics.