A Compare-Aggregate Model with Latent Clustering for Answer Selection

30 May 2019Seunghyun YoonFranck DernoncourtDoo Soon KimTrung BuiKyomin Jung

In this paper, we propose a novel method for a sentence-level answer-selection task that is a fundamental problem in natural language processing. First, we explore the effect of additional information by adopting a pretrained language model to compute the vector representation of the input text and by applying transfer learning from a large-scale corpus... (read more)

PDF Abstract

Evaluation Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK COMPARE
Question Answering TrecQA Comp-Clip + LM + LC MAP 0.868 # 2
Question Answering TrecQA Comp-Clip + LM + LC MRR 0.928 # 2
Question Answering WikiQA Comp-Clip + LM + LC MAP 0.764 # 2
Question Answering WikiQA Comp-Clip + LM + LC MRR 0.784 # 2