The University of Luxembourg, post-trade services provider Clearstream and consulting firm Escent have entered into a research partnership to simplify process anaylsis for IT projects in the financial services industry.
The four-year project will explore how artificial intelligence (AI) and natural language processing (NLP) can make IT processes more efficient.
Within IT projects, stakeholders must define user requirements clearly before determining the needed capabilities and characteristics of a given system.
Professor Lionel Briand, vice-director of the university's Interdisciplinary Centre for Security, Reliability and Trust (SnT), said strict legal and regulatory provisions within the financial industry, as well as the high number of stakeholders involved, made this an especially "laborious task".
"Nearly 50% of budget overruns in IT projects are caused by inadequate requirements that ripple through system design and deployment," he said.
Briand said the project would introduce automation to the 'requirements' engineering process – bringing "incompleteness, inconsistency and ambiguity to a minimum" – and making the analysis of system compliance easier.
Requirements typically contain content written in software modelling and natural descriptive languages.
The team hopes to develop software that will help analysts manipulate and detect inconsistencies between both types of content consistently.
It will also create tools to collect 'acceptance criteria' – the 'pass or fail' conditions a system must meet before it is accepted by a client – from these requirements documents.