Ethics Detector for Software Requirements Documents

Source Code for a paper with the following abstract:

As the use of artificial intelligence grows in both reach and capability, there is an expanding need for ethical considerations at every step of the software engineering design process. The development of sophisticated computational methods for detecting ethical and unethical content in software engineering requirements documents would help ensure these ethical standards are met in a more efficient manner than complete manual review.
Natural language processing (NLP) is the field of techniques that allow computers to read, understand, and formulate text comparable to humans. 
In this work, a novel methodology is proposed that leverages semantics of the words and quantifies ethics in a score.
We present experimental results that elucidate the proposed methodology on a publicly available natural language requirement dataset.