Keynote Talks

Alessio Ferrari

Alessio Ferrari is a Research Scientist at CNR-ISTI (Consiglio Nazionale delle Ricerche-Istituto di Scienza e Tecnologia dell'Informazione "A. Faedo", Pisa, Italy - http://www.isti.cnr.it). His fields of expertise and interest include applications of natural language processing (NLP) techniques to requirements engineering (RE); user and customer interviews in RE; RE education and training; empirical formal methods and empirical software engineering. He is a part of the EU DESIRA project on sustainability, and the EU 4SECURail project on formal methods for railways. He has recently been WP leader of the European Project ASTRail, funded by the Shift2Rail Programme, and participated in other EU projects, such as Learn PAd, about business process models applied to public administration procedures.

Title of the Talk: Ambiguity in Requirements Documents and Elicitation Interviews: NLP Techniques and Beyond

Abstract: Every software project is born with an idea aimed at solving a problem of the real world. The idea is usually expressed by a client and communicated to an analyst during an interview. The analyst will then take care of refining and formalizing the idea in a document of requirements to be met by the software system. The document must be approved by the customer and understood by the developers who will take care of implementing the system. This process of communication, verbal first, and then written, normally occurs through natural language (e.g., English, Italian), which is inherently ambiguous. Ambiguity can generate problems both in the interview phase, and in the implementation phase. To solve this problem, conceptual models have been defined to explain the ambiguity, and automatic NLP solutions have been developed to identify it.

The keynote shows how the problem of ambiguity manifests itself in the interview phase, and within requirements documents. It illustrates NLP techniques that have been developed to detect ambiguities, and will outline how the analysis of speech can also help tackling ambiguity issues in requirements elicitation interviews.