Character-Level Neural Translation for Multilingual Media Monitoring in the SUMMA Project

TL;DR

This paper introduces a novel character-level neural translation approach that jointly addresses story segmentation and clustering in multilingual media monitoring, leveraging low-dimensional semantic representations from sequence-to-sequence models.

Contribution

It proposes a new character-level neural translation model with sliding-window attention for multilingual, morphologically rich languages, enabling joint story segmentation and clustering.

Findings

Effective story segmentation from ASR transcripts.

Successful clustering of stories across languages.

Novel use of translation model representations for media analysis.

Abstract

The paper steps outside the comfort-zone of the traditional NLP tasks like automatic speech recognition (ASR) and machine translation (MT) to addresses two novel problems arising in the automated multilingual news monitoring: segmentation of the TV and radio program ASR transcripts into individual stories, and clustering of the individual stories coming from various sources and languages into storylines. Storyline clustering of stories covering the same events is an essential task for inquisitorial media monitoring. We address these two problems jointly by engaging the low-dimensional semantic representation capabilities of the sequence to sequence neural translation models. To enable joint multi-task learning for multilingual neural translation of morphologically rich languages we replace the attention mechanism with the sliding-window mechanism and operate the sequence to sequence neural translation model on the character-level rather than on the word-level. The story segmentation and storyline clustering problem is tackled by examining the low-dimensional vectors produced as a side-product of the neural translation process. The results of this paper describe a novel approach to the automatic story segmentation and storyline clustering problem.