# A Multi-Task Architecture on Relevance-based Neural Query Translation

## TL;DR

This paper introduces a multi-task learning approach for neural query translation that improves relevance and precision in cross-lingual information retrieval by incorporating a relevance-based auxiliary task.

## Contribution

It proposes a novel multi-task architecture that integrates relevance information into neural machine translation for search queries, enhancing translation quality for IR tasks.

## Key findings

- Achieved 16% improvement over baseline NMT in Italian-English query translation.
- Generated more balanced and precise translations with multi-task learning.
- Demonstrated both quantitative and qualitative benefits of the approach.

## Abstract

We describe a multi-task learning approach to train a Neural Machine Translation (NMT) model with a Relevance-based Auxiliary Task (RAT) for search query translation. The translation process for Cross-lingual Information Retrieval (CLIR) task is usually treated as a black box and it is performed as an independent step. However, an NMT model trained on sentence-level parallel data is not aware of the vocabulary distribution of the retrieval corpus. We address this problem with our multi-task learning architecture that achieves 16% improvement over a strong NMT baseline on Italian-English query-document dataset. We show using both quantitative and qualitative analysis that our model generates balanced and precise translations with the regularization effect it achieves from multi-task learning paradigm.