vector space information retrieval model

This text query system uses vector vector space model for indexing extracted tokens from documents.

Indexing

Tokens will be parsed through each document text and after some cleaning data(punctuation, white spaces), will be index based on vector space model concepts. Tokenized word are stored in a dictionary with information about their repetition and associated document files. Finally tokens are inserted into another dictionary where their weight on TF-IDF scale are assessed. This data are ready to be any query process.

Query

After parsing the documents, program would ask you to enter a query up to 4 words. Entered query will be processed in the same way that document were parsed. Final result is ranked related documents which are assessed by Cosine Similarity formula based on TF-IDF values.