With growing explosion of multi-lingual data on the Internet and other informational and communicational fields, the requirement of having effective automated language identifiers has increased further. More information finds its way into the computer systems and the web and using manual methods to categorize the information is becoming increasingly in-feasible. In this study we discuss improvements we have achieved in existing language identification methods. Couple of new areas that were not explored before is the inclusion of non-Roman scripts and active usage of Unicode information about scripts to enhance the language detection process.
展开▼
