Finnish Native Language Identification

Finnish Native Language Identification Shervin Malmasi Centre for Language Technology Macquarie University Sydney, NSW, Australia shervin.malmasi@mq.edu.au Mark Dras Centre for Language Technology Macquarie University Sydney, NSW, Australia mark.dras@mq.edu.au Abstract We outline the first application of Native Language Identification (NLI) to Finnish learner data. NLI is the task of predicting an author’s first language using writings in an acquired language. Using data from a new learner corpus of Finnish — a lan- guage typology quite different from others previously investigated, with its morpho- logical richness potentially causing diffi- culties — we show that a combination of three feature types is useful for this task. Our system achieves an accuracy of 70% against a baseline of 20% for predicting an author’s L1. Using the same features we can also distinguish non-native writings with an accuracy of 97%. This method- ology can be useful for studying language transfer effects, developing teaching mate- rials tailored to students’ native language and also forensic linguistics. 1 Introduction It has been noted in the linguistics literature since the 1950s that speakers of particular languages have characteristic production patterns when writ- ing in a second language. This language transfer phenomenon has been investigated independently in a number of fields from different perspectives, including qualitative research in Second Language Acquisition (SLA) and more recently though pre- dictive computational models in NLP (Jarvis and Crossley, 2012). Such analyses have traditionally been con- ducted manually by researchers, and the issues that arise when they are attempted on large cor- pora are well known (Ellis, 2008). Recently, re- searchers have noted that NLP has the tools to use large amounts of data to automate this analysis, using complex feature types. This has motivated studies in Native Language Identification (NLI), a subtype of text classification where the goal is to determine the native language (L1) of an author using texts they have written in a second language or L2 (Tetreault et al., 2013). Most work in SLA, NLI and NLP for that mat- ter has dealt with English. This is largely due to the fact that since World War II, the world has witnessed the ascendancy of English as its lingua franca. While English is the native language of over 400 million people in the U.S., U.K. and the Commonwealth, there are also over a billion peo- ple who speak English as their second or foreign language (Guo and Beckett, 2007). This has cre- ated a global environment where learning multiple languages is not exceptional and this has fueled the growing research into language acquisition. However, while English is one of the most prevalent languages in the world there are still a sizeable number of jobs and activities in parts of the world where the acquisition of a language other than English is a necessity. One such example is Finland, where due to the predicted labour shortage, the government has adopted policies encouraging economic and work- related migration (Ministry of Labour, 2006), with an emphasis on the role of the education system. Aiding …

Content truncated to 500 words