Coarse-grained Cross-lingual Alignment of Comparable Texts with Topic Models and Encyclopedic Knowledge

TL;DR

This paper introduces a novel method for aligning comparable texts across languages by leveraging multilingual topics, concept annotations, and theme ordering models, achieving state-of-the-art results in monolingual and cross-lingual alignment tasks.

Contribution

It proposes a combined approach using two-level LDA, HMM, and concept annotations for effective coarse-grained cross-lingual text alignment, which is a new integration in this domain.

Findings

Achieved state-of-the-art performance in monolingual alignment.

Demonstrated effective cross-lingual alignment between English and French.

Validated the method on previously used datasets with strong results.

Abstract

We present a method for coarse-grained cross-lingual alignment of comparable texts: segments consisting of contiguous paragraphs that discuss the same theme (e.g. history, economy) are aligned based on induced multilingual topics. The method combines three ideas: a two-level LDA model that filters out words that do not convey themes, an HMM that models the ordering of themes in the collection of documents, and language-independent concept annotations to serve as a cross-language bridge and to strengthen the connection between paragraphs in the same segment through concept relations. The method is evaluated on English and French data previously used for monolingual alignment. The results show state-of-the-art performance in both monolingual and cross-lingual settings.