A UML Profile for Prediction of Significant Software Requirements
| Autor: | Muhammad Waseem Anwar, Farooque Azam, Ayesha Tariq, Bilal Maqbool, Haider Ali Javaid |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
UML tool
Requirements engineering business.industry Computer science 05 social sciences 050301 education Bayesian network 020207 software engineering Context (language use) 02 engineering and technology Schedule (project management) computer.software_genre Software Systems development life cycle 0202 electrical engineering electronic engineering information engineering Software requirements Software engineering business 0503 education computer |
| Zdroj: | 2019 IEEE 10th Annual Information Technology, Electronics and Mobile Communication Conference (IEMCON). |
| DOI: | 10.1109/iemcon.2019.8936227 |
| Popis: | The preliminary phase of the software development life cycle is Requirements engineering that is nearest to the user’s world. This phase contains tasks that are knowledge concentrated. Therefore, the practice of Bayesian Belief Network (BBN) for modelling this knowledge would be worthful assistance. Accordingly, predicting significant requirements is essential. When poorly identified, numerous problems happen, such as budget overrun, software failures and schedule overrun. However, this phase is usually not performed and skipped by assuming as an inconsequential phase. Significant requirements identification become more vital and challenging when the complexity of the software increases. Therefore, managers and developers recurrently lose confidence in software artifacts. Wavering in the developer’s confidence may in return distress the decisions of which requirement is to implement first. In this paper Bayesian belief network (BBN) approach for predicting significant requirements is proposed to solve the problem by UML Profile mechanism. This probabilistic model supports in accomplishing the uncertainties and inaccuracy generally exists in the requirements engineering process. It makes the prediction more precise, intuitionistic and reasonable. The Profile is imported into a UML tool, which helps in prompt validation of meta-model concepts in practice. The approach is practicable in a realistic context and addresses uncertainties. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|