The work presents “Anwesha: A tool for Semantic Search in Bangla”. This work shows explorations toward building a search engine prototype in Bangla language. The work used sources from WordNet, Wikipedia and statistical co-occurences(LSA) for retrieval of semantically related documents.

For any given query, an Information Retrieval (IR) system is used to obtain and rank relevant word documents from the data collection of interest. The most basic IR system uses Term Frequency Inverse Document Frequency (TF-IDF) to represent documents and queries as vectors, and then uses measures like cosine similarity to assess the relevance of…