# A Compare-Aggregate Model with Latent Clustering for Answer Selection

## TL;DR

This paper introduces a novel answer selection model that leverages pretrained language models and a latent clustering technique, achieving state-of-the-art results on benchmark datasets.

## Contribution

It proposes a compare-aggregate model with latent clustering and transfer learning, improving answer selection performance in NLP tasks.

## Key findings

- Achieved state-of-the-art performance on WikiQA and TREC-QA datasets.
- Demonstrated the effectiveness of latent clustering in answer selection.
- Showed benefits of transfer learning with pretrained language models.

## Abstract

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. Second, we enhance the compare-aggregate model by proposing a novel latent clustering method to compute additional information within the target corpus and by changing the objective function from listwise to pointwise. To evaluate the performance of the proposed approaches, experiments are performed with the WikiQA and TREC-QA datasets. The empirical results demonstrate the superiority of our proposed approach, which achieve state-of-the-art performance for both datasets.