Information_Retrieval_and_Web_Search

TF-IDF index construction, query split computation, named entity disambiguation / named entity linking

Project Part1

• Calculating TF-IDF for tokens and entities in the documents.
• Split queries into different combinations of tokens and entities.
• Calculate the query score for different combinations in order to select the one with highest score.
• Check the Jupyter file inside to see the specification for part 1.

Project Part2

• Use XGBoost with feature selection to build a model for Named Entity Disambiguation, applying TF-IDF and other NLP methods.
• Train the model in Train.py and test the accuracy with test.py.
• Check the Jupyter file inside to see the specification for part 2.