Word and phrase prediction tool for English and Hindi language
| Autor: | Shashi Pal Singh, Yasha Jadwani, Ajai Kumar, Daya Chand Mandad |
|---|---|
| Rok vydání: | 2016 |
| Předmět: |
Hindi
030506 rehabilitation Vocabulary Phrase Helping hand Frequency of occurrence Computer science business.industry media_common.quotation_subject Speech recognition computer.software_genre Variable length language.human_language 03 medical and health sciences Prediction algorithms 0302 clinical medicine language Artificial intelligence 0305 other medical science business computer 030217 neurology & neurosurgery Natural language processing Word (computer architecture) media_common |
| Zdroj: | 2016 International Conference on Electrical, Electronics, and Optimization Techniques (ICEEOT). |
| DOI: | 10.1109/iceeot.2016.7754930 |
| Popis: | Prediction of text while writing has been of utmost importance to people. Prediction has served as a helping hand to people who are slow in typing, or who do not have a good vocabulary. In a country of 336 million populations where people are more frequent in Hindi language, and want to get prediction in Hindi as well. Writing the entire phrases on their own, makes it more prone to errors and mistakes that are made while typing. The prediction tool reduces the chances of having errors while typing. It helps the user for typing efficiently and with a faster speed. The tool is able to predict using multiple algorithms and also we tried to build our own bilingual database which contains phrases of variable length along with their frequency of occurrence in corpus. Whenever a word or phrase is selected from the prediction list, its frequency is increased by one. Consequently, the most frequently used word will have highest frequency/probability leading to increase in efficiency of predictions. Those words and phrases that are not present in the database but have been typed by the user are fed into the database for being used in future predictions. Thus, longer the tool is used, better results it will show Thus, this paper proposes a bilingual word prediction tool that predicts words and phrases for both English and Hindi language. The tool uses different models for the prediction and hence provides a comparison as to which model is better to use and gives better results. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|