Language Segmentation
This addresses language segmentation for low-resource languages, but it is incremental as it builds on existing methods with a new weakly supervised approach.
The paper tackled language segmentation for low-resource languages by comparing supervised, unsupervised, and weakly supervised methods, finding that the weakly supervised approach performed well on short and difficult texts, nearly matching supervised results.
Language segmentation consists in finding the boundaries where one language ends and another language begins in a text written in more than one language. This is important for all natural language processing tasks. The problem can be solved by training language models on language data. However, in the case of low- or no-resource languages, this is problematic. I therefore investigate whether unsupervised methods perform better than supervised methods when it is difficult or impossible to train supervised approaches. A special focus is given to difficult texts, i.e. texts that are rather short (one sentence), containing abbreviations, low-resource languages and non-standard language. I compare three approaches: supervised n-gram language models, unsupervised clustering and weakly supervised n-gram language model induction. I devised the weakly supervised approach in order to deal with difficult text specifically. In order to test the approach, I compiled a small corpus of different text types, ranging from one-sentence texts to texts of about 300 words. The weakly supervised language model induction approach works well on short and difficult texts, outperforming the clustering algorithm and reaching scores in the vicinity of the supervised approach. The results look promising, but there is room for improvement and a more thorough investigation should be undertaken.