Encoding Words into String Vectors for Word Categorization

In this research, we propose the string vector based K Nearest Neighbor as the approach to the word categorization. In the previous works on the text categorization, it was successful to encode texts into string vectors, by preventing the demerits from encoding them into numerical vectors; it provides the motivation for doing this research. In this research, we encode words into string vectors instead of texts, define the semantic operation between the string vectors, and modify the K Nearest Neighbor into the string vector based version as the approach to the word categorization. As the benefits from this research, we expect the improved performance by avoiding problems in encoding texts or words into numerical vectors and more compact representations than numerical vectors. Hence, the goal of this research is to implement the word categorization system with its better performance and more compact representation of words.