A Search Engine Project Incorporating Query Expansion, Classification, Clustering, and Link Analysis.
- Use Boolean Search, TF-IDF, FastText, and Transformer for search
- A search engine with query expansion
- Use Naive-Bayes and fine-tuned Transformer for classification
- Use K-means for clustering
- Page Rank and HITS algorithm for link anlysis
DEMO.mov
- install vue/cli for front project
npm install -g @vue/cli
- install python >= 3.8 with following guide for back project
- clone the project with following command
git clone [email protected]:IR1401-Spring-Final-Projects/News1401-13.git
- install requirments with following command
pip install -r requirements.txt
- some models are not on git due to maximum file size so for running project download them from links given and locate them in given path
- path: data/news.json | link: link
- path: models/classification_logistic_regression_improved/fasttext/fasttext.bin | link: link
- path: models/classification_transformers/pytorch_model.bin | link: link
- path: models/clustering/fasttext/fasttext.bin | link: link
- path: models/fasttext_search/fasttext.bin | link: link
- path: models/preprocessed_data/data.plk | link: link
- path: models/preprocessed_data/clf_data.plk | link: link
- path: models/QE_fasttext/fasttext.bin | link: link
- path: models/QE_fasttext/all_words_vectors_emb_fasttext.json | link: link
- path: models/transformers_search/transformer_model.model/pytorch_model.bin | link: link
- path: models/transformers_search/transformer_vectors_emb.json | link: link
- run backend project with following command
uvicorn src.api:app --reload
- go to
front/search_engine
and run frontend project with following command
npm run serve
enjoy it!