Modelling and Managing Inconsistent Software Requirements



Brief Description

Objectives:

The aims of this project are to address the problem of handling inconsistent requirements in the requirements stage. We are especially concerned with inconsistency, redundancy, vagueness, and incompleteness. We focus on a unified logical tool to represent requirements specifications and to characterize the inconsistent requirements. The main features of the proposed framework for handling inconsistent requirements are: (A) To represent both the clear and vague requirements; (B) To characterize inconsistent requirements; (C) To provide flexible strategies for handling inconsistent requirements based on Annotated Predicate Calculus. The main objectives of this research are: (1) to develop an unified logical tool for formalizing requirements statements; (2) to provide a clear formal characterization of inconsistent requirements; (3) to develop a series of flexible strategies for handling inconsistency.

Achievement/findings as the result of the award:

1. We have presented priority-based approaches to measuring the inconsistency in requirements specifications, which may help developers and stakeholders resolve inconsistency and make some necessary trade-off decisions.

2. We have investigated a negotiation-based approach to managing software requirements changes, which may help developers and stakeholders resolve inconsistency and make some necessary trade-off decisions when new requirements are obtained or when some requirements (which were removed earlier) have to be added .

3. We have provided a logic-based framework for handling Non-canonical software requirements:

  • Constructed a kind of Annotated Predicate Calculus appropriate for formulating vague requirements as well as clear requirements;

  • Characterized the non-canonical software requirements based on Annotated Predicate Calculus, including classical inconsistency, vague inconsistency, potential inconsistency, incompleteness and redundancy;

  • Presented an inconsistency-based strategy for clarifying vague requirements;

  • Presented strategies to identifying and analyzing redundancy.

    4. We have presented a vote-based approach to identifying acceptable common proposals for handling inconsistency that results from conflicting intentions of different stakeholders during requirements stage.

    Other details

  • Period: 2006- 2008
  • Type: Research Project
  • Status: Completed
  • Funding Body: British Royal Society and Chinese Academy of Sciences, £12,000.00

    Personnel Involved (UK)

  • Prof David Bell
  • Dr. Weiru Liu
  • Dr. Jun Hong

    Personnel Involved (China)

  • Prof Ruqian Lu
  • Prof. Zhi Jin
  • Dr. Kedian Mu