Using natural language processng and answer set solving techniques this project reads in multiple tickers and by information extraction and merging them in an intelligent way we output a comprehensive summary of the given tickers.
python2 run.py [-h] [--verbose VERBOSE] [--kicktionary KICKTIONARY]
[--verbnet VERBNET] [--tickers TICKERS] [--luorder LUORDER]
optional arguments:
-h, --help show this help message and exit
--verbose VERBOSE print helpful messages about the progress
--kicktionary KICKTIONARY
location of kicktionary xml file
--verbnet VERBNET location of folder with verbnet xml files
--tickers TICKERS location of tickers folder (by default it's "data/input", so you can put tickers inside)
--luorder LUORDER location of folder with lexical unit order file
project_folder
├── run.py
├── data
| └── ...
├── parser
│ ├── anna-3.3.jar
│ ├── models.de
│ │ ├── lemmatizer.model
│ │ ├── mtag.model
│ │ ├── parser.model
│ │ └── tagger.model
│ ├── models.en
│ │ ├── lemmatizer.model
│ │ ├── parser.model
│ │ └── tagger.model
│ ├── parse.sh
├── extracting
| └── ...
├── preprocessing
| └── ...
├── reasoning
└── ...
All ticker files need to be placed in the same directory. The directory is easily selected using the commandline argument -- tickers DIR
.
The results are stored in the data/output folder. Each run gets its own subdir named by a timestamp.