There can be no debate that Information Mining Projects cause processing tools to arise with the sole purpose of converting available organization data into useful knowledge on account of decision-making. Considering the aforementioned, this type of projects demands due diligence in the requirement specification process, as the latter needs to be thorough and traceable throughout the entire project, and therefore, process associated with requirements engineering are not to be reused in any future projects among the like. Similarly, latest methodologies within the field of Information Mining fail to take practices associated to stakeholders and costumers' requirements management into account. With this aim in mind, a solution model to the Information Mining Project Management necessities is proposed. • Computing methodologies➝Modeling and simulation • Information systems➝Information systems applications
The basis of Information Mining is centered on data processing for
nontrivial knowledge collection, an organizational task in which
analysis and synthesis tools are indispensable [
All things considered, this paper employs a proposed model which contributes to the development of a thorough requirements management in the context of an Information Mining project. In order to do so, the detected problem is first described (Section 2), and then a solution proposal is suggested (Section 3). Afterwards, a case study in which the proposed process is implemented is then shown (Section 4), and lastly, conclusions and future work lines are presented (Section 5).
What a proper Information Mining project needs is a due
consistent and traceable the project throughout - requirements
specification, which should, at first, allow for an orderly project
management, and also leave out any possible requirements
misunderstandings [
At the origination of an Information Mining project, its objectives
describing customer’s general necessities - what he wants to get as
a final result of the project, generally linked to strategic and tactical
business goals [
Latest methodologies - CRISP-D, P3TQ and SEMMA - are
centered on Knowledge Discovery in Databases Process (KDD)
and emphasize available data detection, together with a simplified
overall vision of where the project develops [
A proposed process model is split into five orderly phases: “Project
Definition”, “Business Process Engineering”, “Business Process
Data Engineering”, “Business Conceptualization” and
“Information Mining Process Specification”.
“Project Definition” phase aims to define the project scope,
stakeholders, and objectives to accomplish. The “Business Process
Engineering” phase seeks to identify and survey the most relevant
business processes in the project. The purpose of "Business Process
Engineering” is to locate data repositories where information of the
various business processes is stored and to survey their contents.
The phase of "Business Conceptualization", attempts to define
business in terms of concepts developed and vocabulary managed
in order to understand the business jargon, therefore revealing
business technical words' meanings coined in the business context.
Finally, "Information Mining Process Specification" phase intends
to identify Information Mining Processes available for use in
solving business processes problems, prior to developing the
planification of the remaining project activities. A description of
the proposed model phases and activities is displayed at [
In this section, the results obtained applying the proposed process
model to a case study, are offered. Such case study uncovers
behavior patterns allowing the description of trailers used as
resources in the production area of a metallurgical enterprise. These
patterns will be taken under consideration for decision making
when planning the assembly line of units. Following, the products
obtained in the five phases of the process are shown (for clarity
generated graphics and application models of each phase they are
also represented in [
The following activities are described: “Identifying Project Objectives” (Figure 1), “Identifying Project Stakeholders” (Figure 2), and “Identifying the Project Scope” (Figure 3).
1. Documentation analysis to detect project objectives. 2. Generation of a functional decomposition tree (figure 4). 3. Correspondence between function decomposition tree and objective documentation templates (figure 5).
Output: Project and Requirement Objectives Templates (figure 6).
The following activities are described: “Identifying Business Processes” (figure 17) and “Surveying Business Processes” (figure 18).
Input: Project Scope Definition Template, Project Objective Template, Requirement Objective Template, Project Success Criteria Template and Project Expectations Template.
Process: 1. Documentation analysis to define business activities related to the project objectives and generation of the business processes diagram template. 2. Generation of the column of business process, associated with each stakeholder in the project stakeholder template (defined in the first phase of the process).
1. Conduction of interviews with stakeholders to understand each business process and to generate a business process template per process.
Output: Business Process Templates. The business processes defined are two: “Units Production” and “Production Planning” (figure 20).
The following activities are described: “Identifying Data Repositories” (figure 21) and “Surveying Data Repositories” (figure 22).
Input: Business Process Diagram Template, templates corresponding to every business process and Stakeholders interviews information. Process: 1. Documentation analysis to define the data repositories used or consulted by each business process and generation of the data repositories template.
The following activities are described: “Developing a Business Dictionary” (figure 25) and “Developing a Business Model” (figure 26).
Process: 1. Documentation analysis to define the business keywords.
Upon identification, words are validated by stakeholders and thus the business dictionary is built up. 2. Wording glossary build-up, detailing keywords. By means of example, in figure 27 two keywords of the utmost importance to the business domain. 3. Building of the Concept-Relation and Concept-AttributesValue chart (figures 28 and 29, respectively) upon the wording glossary, and generation of the Concept relations graph from the two latter charts (figure 30). 4. Correspondence establishment between wording glossary and definitions, acronyms and abbreviations template (figure 31). 5. Correspondence establishment among concept-attributevalue chart, wording glossary, and requirement-related attributes template (figure 32).
Output: Business Dictionary Template, Template of Abbreviations, Acronyms, and Definitions and Template of Attributes related to Requirements. Figure 33 shows an extract of each template.
Input: Data Repositories Template, Data Structure Template and Business Dictionary Template.
Process: 1. Documentation analysis to establish the relationships among business processes, data repositories and business words and business model diagram build-up.
The following activities are described: “Formalizing Business Problems” (figure 35), “Identifying Information Mining Processes” (figure 36) and “Developing a Project Plan” (figure 37).
Input: Project Scope Definition Template, Project Objective Template, Requirement Objective Template and Business Process Diagram. Process: 1. Analysis of the project objectives, project scope definition, business processes and data repositories and development of a business problems list. 2. If necessary, project objectives and requirement templates are verified and updated in parallel to the former process.
This paper's main contribution is to provide the community with tools so as to enable them to carry out a thorough management of requirements in the context of an Information Mining Project. Consequently, this paper seeks to improve current methodologies in which this approach has been severely overlooked, which is why the process model proposed was applied to controversial casestudies, covering various business issues. This being the case, both customer requirements and requirements-related project objectives were successfully and efficiently managed, as well as information regarding business processes and their problems, together with the data repositories employed, were obtained. Hence, the development of a project master plan - guiding the project through its very stages - was successfully developed.
As future lines of work, the in-depth analysis and description of structural procedures is heavily suggested, as it is believed that the statement of every document process can lead to a higher rate of automation in these processes' techniques. Moreover, it is considered that a software tool capable of assessing the condition of templates and minutes advised by the process - additionally distinguishing the various versions of a single document, thereby reporting each version's contents together with a change history - is highly desirable project to be substantiated.
The research presented in this paper has been partially financed by the 11211 EIU TIBA, research project of the National University of Technology - UTN (for its acronym in Spanish) Buenos Aires.