ZhihuRec Data-mining

A flask app for analyzing ZhihuRec dataset.

Requirement

 pip install requirements.txt

Usage

• [Dataset] Put dataset ZhihuRec in the root directory.
• [Work Path] Set the work path in root directory.
• [Preprocess] Run the io.py, to convert answer_infos.txt into .csv files.

1.First, run this command to get answers' csv files:

 python tools/io.py

Or just download from here:

Baidu NetDisk 
Link:https://pan.baidu.com/s/1Ey-R9yo6_HNuoZuhEJivjg 
Code: 8rc7

Unzip and put the folder answer_csv into source/

2.Then you can use this command to run the flask app:

 python app.py

The flask app will run on the "127.0.0.1:5000"

Files

• [model] The tf-idf model will be saved here.
• [source] Processed files
  • [answer_csv] Answers' csv files. All files are sorted.
    • [xxxx.csv] The xxxx means the start(min) answer's index in this file.
• [tools] Tools help you analyze the dataset.
  • [io.py] Used to read/write/convert dataset.
  • [tfidf.py] TF-IDF algorithm. its mainly functions are
    • train()
    • load_tfidf()
    • save_tfidf()
    • compare_similarity().
• [zhihuRec] The dataset. You should put txt files here.
• [app.py] The entry of the flask app.
• [preprocess.py] Use the code in tools to create tfidf matrix, and save the result into model.