IRHomeWork

Big homework for Information Retrieval.

Dependency

• Python: 3.7
• Flask: 1.1.1
• Elasticsearch: 7.1.0
• Elasticsearch-full: 7.5

How To RUN?

1. Start elasticsearch server

/usr/local/bin/elasticsearch

2. Edit config

edit config.py

maps # index setting
files_to_handle # source file of corpus. Be preprocessed by thulac

3. Build index

python3 build_index.py

4. Run application

After build index, run application

WARN: Use the same index name in app.py and build_index.py

python3 app.py

Then open http://localhost:5000 in your browser. You can see following page.

5. (Optional) Build word vector

   build word vector to optimize result

   • Use word2vec.py to get word vector

   python3 word2vec.py
   --iter: 迭代轮数,int
   --max_doc：使用多少句子,int OR None
   --mode：train/count/test中的一个,str
   --model：存放词向量的路径,str
   --doc_print：每隔多少个句子输出一次,int
   --min_cnt：训练词向量时的最小词频,int
   --workers：多线程中使用多少个线程,int
   --step_store：每隔多少步存储一次词向量,int