Abstract Millions of software are lunched yearly and this software depend on data to produce required output. Personal data privacy and security has been a source of public concern for some time, and is usually interpreted in terms of data obtained from interaction with software. It is difficult to know whether a software system's decisions are fair and what considerations were put in place in the system's internal decision-making process if the system's decisions are opaque. This has the potential to cause injustice and bias. In addition, a lack of openness may lead to a decrease in user acceptance and happiness. Algorithmic data-driven decision-making systems are becoming more automated, and they've had a lot of success in a lot of different applications. The General Data Protection Regulation of the European Union and other regulations limits algorithmic use of personal data and has fueled the dispute over the right to disclosure. This research adapted a crowd requirements elicitation model to develop a framework for the proper elicitation non-functional requirement. The developed model uses natural language processing integrated into a chatbot and a document extraction strategy since non-functional requirement exist also as government regulations and industrial standards. Proper and comprehensive elicitation of non-functional requirements will give accurate information on how the system performs its required task and such documents are best in terms of openness to the use of data by algorithms to avoid algorithm decision making concerns.
原文链接
分享
引用
