lexi_sent

A python3 script for lexicon-based sentiment analysis.

Usage

The program runs from the command line. Input can be a csv file or a folder.

To see the program help:

python3 lexi_sent.py --help 

For CSV import, run the following commands:

python3 lexi_sent.py test.csv 

The input should be a CSV file with a single unnamed column, containing a document per row (see test.csv).

The folder import expects the input to be a folder. Only txt files can be imported. Example call:

python3 lexi_sent.py test

Optional arguments:

• -o to specify a custom output file (defaults to 'Sentiments.csv')
• -w to specify a custom wordlist. Defaults to 'MPQA.csv'. Custom wordlists need to be in the same format as the default wordlist.
• -v prints name of each file during folder import, for tracing of encoding or other issues.

FAQ

How to install dependencies?

If you're getting the error "ImportError: No module named ...", install the missing package(s):

pip3 install pandas numpy openpyxl

How to install Python?

If you're on Mac or Linux, Python is already installed. On windows, get the latest version from the Python website

Citation

For usage in any published material, please cite:

Nauhaus, S. (2020). The Role of Stakeholder Sentiment in Strategic Decision-Making: A Behavioral Perspective. Geneva: University of Geneva.

If you're using the default MPQA dictionary, the authors would appreciate a citation as well:

Theresa Wilson, Janyce Wiebe, & Paul Hoffmann (2005). Recognizing Contextual Polarity in Phrase-Level Sentiment Analysis. Proc. of HLT-EMNLP-2005.