The paper concerns effort estimation of software development projects, in particular, at the level of product delivery stages. It proposes a new approach to model project data to support expert-supervised analogy-based effort estimation. The data is modeled using Semantic Web technologies, such as Resource Description Framework (RDF) and Ontology Language for the Web (OWL). Moreover, in the paper, we define a method of supervised case-based reasoning. The method enables to search for similar projects’ tasks at different levels of abstraction. For instance, instead of searching for a task performed by a specific person, one could look for tasks performed by people with similar capabilities. The proposed method relies on ontology that defines the core concepts and relationships. However, it is possible to introduce new classes and relationships, without the need of altering the search mechanisms. Finally, we implemented a prototype tool that was used to preliminary validate the proposed approach. We observed that the proposed approach could potentially help experts in estimating non-trivial tasks that are often underestimated.
3
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
Most of the use-case-based effort estimation techniques rely on a concept of use-case transaction that is used for measuring complexity of use cases. This includes the most recognizable Karner's Use Case Points and the Transactions metric recently proposed by Robiolo and Orosco. Both approaches are focusing on quantitative aspects of use-case transactions. However, in this paper, we would like to propose a new method for measuring use-case complexity that is called TTPoints. It includes knowledge regarding semantics of transactions, numbers of business objects and interacting actors. Effort estimation accuracy of the proposed method was compared to accuracies of Use Case Points, Transactions, and Entity Objects, in the case study on a set of 8 projects.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.