Vector-Representation-Doc2Vec

This repo shows us how to compute similarity between two documents. Documents can be of any length or in some specifc case, one document can contain a user's query and other documents can be text files from where the query will be asked and Doc2Vec can be used to show the similarity between that query and other documents. The maximum score will point to the document from where the query was initially taken.

1. Extract the file from the squad_train_doc.json folder.
2. Load the JSON file.
3. Run the model and implement the model.
4. Input your question from any of the documents, using list_questions_per_doc[index].
5. Compute the dataframe.
6. Compare the results.