Dear Members, I want to make a corpus of affective norms for Hindi like has been made for English norms. There are many corpus for English words but I didn't find any good resource for Hindi. If anyone has any idea about such good resources please share the information about it.
Our model is capable of handling a variety of characteristics seen in Indian Social Media.The Indian social media mostly use phonetic typing i.e. the word is transliterated to Roman English based on the pronunciation of the word. This word would have different transcription as we are expressing the pronunciation of a word which depends on various factors. One of the major factor, is the mother tongue of the person. The same word is pronounced differently across different states in India. The other factor for phonetic typing is the limitation in the number of English phonetics when compared to any other Indian languages. slight variations we have designed a model, which could understand the variations in the spelling and transliterate into Devanagari script. Our model checks the following equivalences:
1 Homorganic nasal case checks equivalence
2 Anuswar-chandra Bindu exchange
3 Non-obligatory use of Nukta
4 Homophonic ending
5 Canonical Unicode
6 popular usage
Are you familiar with Sketch Engine (www.sketchengine.co.uk)? There is a 60 million corpus of Hindi (HindiWaC, https://www.sketchengine.co.uk/hindiwac-corpus/), maybe you can use this data?
- Badges
- Science topic
- Similar topics
- Linguistics