Predicting the Next Word of a Sinhala Word Series Using Recurrent Neural Networks

Making accurate predictions before occurrence have been a useful phenomenon lately with the advancement of technology. Numerous technologies are used to make countless word predicting applications that make typing easier. Unlike low inflected languages like English, it is not easy to build word predictors for highly inflected languages such as Sinhala language, which comprises a rich set of vocabulary and grammar rules. The aim of this research is to find possible ways to build a word predictor for Sinhala language, by identifying patterns using Recurrent Neural Networks, considering a sequence length of two to predict the next word.