Most of the professional activities are created as projects. Project management is a perplexing process including numerous elements interrelated and reliant on outside agents that may confuse its control. Projects, as ealier defined by PMBOK of Project Management Institute (PMI, 2009), are designed to solve an issue or need, have an impermanent impact and is exceptional in time and not repeatable in the same circumstances. Vulnerability is a key factor associated with project management. This factor affects the consumption of resources, the estimation of time and cash as well as the effect of risk and quality.
Customary project management, regardless of sector, identifies a series of phases such as:
1. Initiation, where needs are distinguished and assessed to know whether it is possible to carry out the project at this stage. The vulnerability is high because of absence of exact data, which means that the possibility of blunder in the estimate is high.
2. Planning, this aims to build up a solution in more noteworthy detail, by separating the issue into more point by point activities. This reduces vulnerability and makes estimates and forecasts. You must also characterize the tasks and date-book and estimate the time and cash expected to attempt the project.
A standout amongst the most vital objectives in the management of data technology projects has been focused on the right estimation of exertion and risk recognition of defects or quality nonconformities in the usage of the system.
Making a model or knowing the agents affecting the risk, as well as their severity, will permit the managers of software projects to encourage project management starting process.
As of late, it has been noticed a pattern to use computational intelligence techniques based on data to project management problems and these techniques were demonstrated ready to distinguish complex relationships.
the phase which will profit more from the execution of data mining techniques is the underlying arranging phase.
Shockingly those data can’t be easily processed physically because of their heterogeneity, and that is the ideal condition for Data Mining, which performs the processing of data extraction from crude data with a specific end goal to remove useful conclusions. Data mining is equipped for examining the data making rules or conclusions giving the association tools committed to decreasing the risks and vulnerability in the decision-making process.
Data mining can be useful in all stages and fields: estimating better costs, streamlining the bids, assessing the risks, decreasing the vulnerability in the span of tasks, and so on.
This need arises to model for exertion and the risk affect measured as defects in the project. System must be considered as a data mining project, for which it has been taken after the phases given by CRISP-DM technique (Chapman et al., 2000). The important steps to be taken after the start by data understanding; this stage is a study of accessible data. The following stage is to play out the data arrangement to modeling, since every method needs a specific sort of data.
After the use of the data mining analysis described, it can be gotten a decrease of vulnerability certain in the projects, especially in data technology systems. What’s more, you can get more results as, for instance, relationships between attributes, distinguishing proof of variables that give data or are more significant to estimate the exertion or to recognize the level of risk o quality assurance.
The quality variable, specific models, can be produced using similar data mining as the one used for time and cost estimation however to anticipate this resulting quality, which on account of software projects it is usually identified with the quantity of defects.
Given the high nonlinearity of the process and its reliance on non-quantifiable parameters, the case study is confined inside the problems in which the use of intelligent data mining techniques has given successful results, with the presentation of significant control advances.
Given the case of study is presented as a data mining application process, it has been considered the use of one of the all the more across the board methodologies: CrossIndustryStandard Process for Data Mining (Chapman et al., 2000) CRISP-DM. This has just been used with a specific end goal to solve similar problems (Montequín et al., 2005).
This technique defines the data mining life cycle process; it consists of 6 phases, this is a worldwide process which is performed by process iterations. Likewise, the phases communicate with each other all through the development process