Hidden fuzzy information: Requirement specification and measurement of project provider performance using the best worst method

The requirement specification process is an important part of a project and has the potential to prevent problems that may last for years after a project is delivered. Previous studies on the requirement specification process have focused on clarifying stated fuzzy terms in software requirement engineering. However, in many projects there is information that is not stated, but it is implied and can be inferred. This hidden information is usually ignored due to the assumption that 'the provider understands what they mean/need'. This assumption is not always true. Such information, if extracted, may include fuzzy terms, namely hidden fuzzy terms (HFTs), which need specification. Therefore, these fuzzy terms have to be identified and then specified to avoid potential future consequences. This study proposes an algorithm to extract the hidden fuzzy terms, utilises a fuzzy inference system (FIS) to specify them, and applies the best worst multi-criteria decision making method (BWM) to evaluate the delivered product and measure the performance of the provider. The model is then used to examine a case from Defence Housing Australia. Such evaluation and measurement enable the project owner/manager to have a transparent basis to support decisions later in different phases of the project, and to ultimately reduce the likelihood of conflict and the receipt of an unsatisfactory product. (C) 2019 Elsevier B.V. All rights reserved.