Using Ontologies and Machine Learning for Hazard Identification and Safety Analysis

Abstract

This book synthesizes the work of the managing requirements knowledge (MARK) community during the last 5 years. The first idea to organize a workshop on this topic came to our minds in winter 2007. We were both working on our Ph.D. projects at the Technische Universität München (TUM) under the supervision of Bernd Brügge. Anil was focusing on software product lines, while Walid was looking at the application of ontologies and machine learning to collaborative software engineering, in particular during bug fixing and API reuse. Our fields of interest seemed divergent at first glance. However, after a couple of discussionsalso with colleagues from industry-we found that some of the problems we were trying to address are very similar. Valuable experiences and knowledge gained in the course of software projects, in particular during the work with requirements, remain tacit in the mind of people. The same problems in understanding and implementing requirements occur again and again. We were convinced about the need for a new perspective on requirementsconsidering them as a knowledge asset in software organizations-in addition to the engineering and lifecycle perspectives. We were convinced about the huge potentials of recent trends such as ontologies, wikis, Web 2.0, recommendation systems, and data mining, to the requirements engineering community. In the last years, the MARK workshop successfully took place in Barcelona, Atlanta, Sydney, and Trento. It has been one of the most successful workshops at the IEEE International Conference on Requirements Engineering that is based on submission and registration statistics, as well as the feedback of the participants. The achievements are remarkable. Novel approaches such as "recommending features and stakeholders by analyzing requirements repositories" or "using semantic wikis to represent and reason about requirements" have found their way to main conferences and journal in the field. Some of the tools are already being used in practice. v

Enhancing Misuse Cases With Risk Assessment for Safety Requirements

“…There are few research studies that had investigated IS safety hazards identification and RE. Olawande et al [16] introduced an easy starter for identifying potential system safety using HazId approach. Authors stated that the proposed method helped in reducing cost of safety analysis through the use of technologies such as case-based reasoning, ontology, and natural language processing.…”

Semi-automated development of conceptual models from natural language text

“…Ontologies support many requirement modelling styles, including textual requirements such as in Chicaiza, López, Piedra, Martínez and Tovar (2010), Daramola, Sindre and Moser (2012) and Daramola, Stålhane, Omoronyia and Sindre (2013). Examples of ontology use with UML include Boukhari, Bellatreche and Jean (2012), Cardei et al (2008) and Castañeda et al (2012).…”

Towards Explainable Direct Marketing in the Telecom Industry Through Hybrid Machine Learning

“…Case-based reasoning (CBR) uses the record of past occurrences to provide the solution to a new occurrence. The approach has been found to have strong explanation mechanisms because it derives its explanations from similar previous cases [ 13 , 14 ]. Typically the CBR problem-solving process entails case retrieval, case reuse, case adaptation, and case retention.…”