Natural language processing in the eyes of software engineers


Natural language (NL) is prevalent in the Requirements Engineering (RE)
field (a major subfield in Software Engineering). RE is embedded within
a diverse organizational context, where engineers may have to interact
with people from different domains and backgrounds such as legal
experts, bank managers, system analysts, software developers, etc. While
using NL creates a common understanding among those stakeholders, it
comes with the cost of having to process a considerable amount of text
spanning over hundreds of pages. To alleviate this problem, engineers
may use NLP tools to extract and categorize useful information
automatically. The question is, then, to what extent are current NLP
off-the-shelf tools usable?

In this talk, I will discuss two different project scopes where
requirements engineers need NLP technologies: (i) early quality
assurance of critical systems in order to avoid serious implications
that might be caused by ambiguous, inconsistent or incomplete
requirements; (ii) readability level and compliance verification of
legal artifacts (e.g., privacy policies) according to international
regulations such as the General Data Protection Regulation (GDPR)
enforced in Europe.

*********

Time: Monday, 26th of August 2019, 6:30 p.m. sharp

Location: Oesterreichisches Forschungsinstitut
fuer Artificial Intelligence, OFAI
Freyung 6, Stiege 6, 1010 Wien

OESTERREICHISCHES FORSCHUNGSINSTITUT
FUER ARTIFICIAL INTELLIGENCE
Univ.-Prof. Ing. Dr. Robert Trappl

*********