Lexical triggers and latent semantic analysis for cross-lingual language model adaptation

Abstract

In-domain texts for estimating statistical language models are not easily found for most languages of the world. We present two techniques to take advantage of in-domain text resources in other languages. First, we extend the notion of lexical triggers, which have been used monolingually for language model adaptation, to the cross-lingual problem, permitting the construction of sharper language models for a target-language document by drawing statistics from related documents in a resource-rich language. Next, we show that cross-lingual latent semantic analysis is similarly capable of extracting useful statistics for language modeling. Neither technique requires explicit translation capabilities between the two languages! We demonstrate significant reductions in both perplexity and word error rate on a Mandarin speech recognition task by using these techniques.

Latent Semantic Analysis and Machine Translation

“…Cross-language information retrieval search results in languages differing from the query has also received attention (Rehder and et al 1998) as has LSA's use in language modeling (Kim and Khudanpur, 2004).…”

Computational Linguistics and Intelligent Text Processing

“…Yet, they do not use a unified framework for both tasks. Some researchers cast uncertainty detection as a token sequence labeling problem, and then use hand-crafted rules to extract scopes [15,44,48,49]. Others define both tasks as token sequence labeling problems; yet use separate feature sets for each tasks or do not use the output of one task to inform the other [50].…”

“…L T -words) that are topically related to s. For example, given that we sampled ('bacteria') from the L S -document, we are very likely to sample words like ('bacteria') or ('disease') from the L T -document. 7 This is similar in spirit to the idea of lexical triggers [48]. We can use a baseline scoring function S raw (as defined in Section 3.1) and define the trigger probability P CC (t|s)…”

Mining monolingual and bilingual corpora

“…Inter-lingual triggers have been also used in [29] to enrich resource deficient languages from those which are considered as potentially important. An inter-lingual trigger is henceforth a set made up of a word (or a sequence of words) f in a source language, and its best correlated words in a target language e 1 , e 2 , .…”