- python3, stanford_openie, vaderSentiment, etc.
pip3 install -r requirements.txt
python3 extract_reviews.py [-h] <arguments>
Example:
python3 extract_reviews.py --data_path ../data/australian_user_reviews.json
--data_path
: Directory path to the data. Default:../data/australian_user_reviews.json
--output_dir
: Directory path to the output. Default:reviews
python3 extract_recommended.py [-h] <arguments>
Example:
python3 extract_recommended.py --data_path ../data/australian_user_reviews.json --lower
--data_path
: Directory path to the data. Default:../data/australian_user_reviews.json
--output_dir
: Directory path to the outputs. Default:reviews
--lower
: Boolean flag activating lowercasing of the reviews output. Default:False
python3 scrap_steam_tags.py [-h] <arguments>
Example:
python3 scrap_steam_tags.py
--input_dir
: Directory path to the reviews. Default:reviews
--output_dir
: Directory path to the figures. Default:figures
python3 vader.py [-h] <arguments>
Example:
python3 vader.py --data_path ../data/australian_user_reviews.json --verbose
--data_path
: Directory path to the data. Default:../data/australian_user_reviews.json
--input_dir
: Directory path to the reviews. Default:reviews
--output_dir
: Directory path to the pickle output. Default:bin
--verbose
: Boolean flag activating console prints. Default:False
python3 sequences.py [-h] <arguments>
Example:
python3 sequences.py --recommend 1 --sentiment pos --topn 200 --write
--input_dir
: Directory path to the dataframe pickle. Default:bin
--output_dir
: Directory path to the output sequences. Default:reviews
--recommend
: Recommendation attribute of reviews. Choices:{0, 1}
. 0 = False, 1 = True. (required) Default:None
--sentiment
: Sentiment connotation of reviews. Choices:{pos, neu, neg}
(required) Default:None
--topn
: Number of the most of common sequences. Default:10
--write
: Boolean flag activating writing file output. Default:False
python3 stanfordopenie.py [-h] <arguments>
Example:
python3 stanfordopenie.py --input_path ./reviews/reviews_recommended_true.txt --interval 0 90000
--input_path
: Input path to a review. (required) Default:None
--output_dir
: Directory path to the triples output. Default:triples
--interval
: Interval of lines to read. (required; 2 arguments) Default:None
Thach Jean-Pierre - University of Montreal
Wong Leo - University of Montreal