Software Project Management activities are classified as planning, monitoring-control and termination. Planning is the most important activity in project management which defines the resources required to complete the project successfully. Software Cost Estimation is the process of predicting the cost and time required to complete the project. The basic input for the software cost estimation is coding size and set of cost drivers, the output is Effort in terms of Person-Months (PM's). In this paper we proposed a model for software cost estimation using Multi Objective (MO) Particle Swarm Optimization. The parameters of model tuned by using MOPSO considering two objectives Mean Absolute Relative Error and Prediction. The COCOMO dataset is considered for testing the model. It was observed that the model gives better results when compared with the standard COCOMO model. It is also observed, when provided with enough classification among training data may give better results. KeywordsKDLOC-thousands of delivered lines of code, PM-person months, PSO-particle swarm optimization, COCOMOconstructive cost estimation, MO-Multi Objective.

Abstract

In the last decades the complexity of software development projects had a significant increase. This complexity emerges from the higher degree of sophistication in the contexts they aim to serve and from the evolution of the functionalities implemented by the applications. The Service Oriented Architecture (SOA), a document review investigation has taken place with the objective to compare SOA development with two highly used agile software development methods: Extreme Programming (XP) and Relational Unified Process (RUP). The motor behind this investigation is to see if this agile development methods are capable enough to design effectively and efficiently SOA type applications. The investigation concluded that both methodologies are not exactly suited to create SOA applications, but RUP has a greater chance to provide SOA type applications if the proper adjustments are made, specially on the design side of the methodology.

No abstract