Dealing with unknowns in machine translation

An 'unknown' is defined as a word for which there is no entry in the dictionary used by the translation system. In general, a text may contain several unknowns. These words may be names, acronyms, abbreviations, terminology or foreign words. It is a common practice in India to mix the words of English in Hindi and other Indian languages and vice-versa. However, the grammatical rules in construction of gender, number, verb-normalization or forms, conform to that for the language used irrespective of their origin. This gives rise to a frequent encounter of unknown words in day-to-day communication. A machine translation system has to provide mechanism for handling such unknowns. Spelling mistakes is yet another source that contributes to the unknowns. In this paper we describe a strategy being adopted in our system for machine aided translation from English to Hindi. No attempt has been made to expand the vocabulary by deriving their meaning. Instead, once an unknown is identified, a transliteration in Hindi with appropriate suffixes or appendage is used to substitute for their meaning. We use predictive parsing and a number of heuristics to identify the type of unknown.