temiedani / nlp_vector-space-semantics

NLP_Vector-Space-Semantics

The aim of the implementation was to create a vector space representation of a document containing lines spoken by a character in the Eastenders TV series script that can distinguish between lines spoken by each character. This distinction is measured by how well a simple information retrieval classification method can select documents from validation and test data as belonging to the correct class of document (i.e., deciding which character spoke the lines by measuring the similarity of those document vectors to those built-in training)