CMP681 - Neural Query Expansion Project

İsmail Talha Yılmaz

N18140956

ExtractTweets.py -> extraction of tweets from JSON file

CleanTweets.py -> pre-processing of extracted tweets

PrepareQueryTweetsBM25.py -> matching tweets with most relevant queries

EncoderDecoderModel.py -> seq2seq architecture and predictions

ExpandQueriesByEmbeddingSimilarity.py -> calculate similar predicted query words to tweet and use them for expansion

Evaluate.py -> Evaluation of overall system

External Sources:

Google News pre-trained embeddings were used

TREC 2011 Microblog dataset is used