Examinando por Materia "CRISP-DM"

Mostrando 1 - 3 de 3
  • No hay miniatura disponible
    Publicación
    Sistema multiagente basado en modelos de lenguaje para optimizar la selección y asignación de talento en outsourcing
    (Universidad EAFIT, 2025) Álvarez-Petro, Oscar David; Martínez Vargas, Juan David
  • No hay miniatura disponible
    Ítem
    Towards an improved ASUM-DM process methodology for cross-disciplinary multi-organization big data & analytics projects
    (Springer Verlag, 2018-01-01) Angée, S.; Lozano-Argel, S.I.; Montoya-Munera, E.N.; Ospina-Arango, J.-D.; Tabares-Betancur, M.S.; Universidad EAFIT. Departamento de Ingeniería de Sistemas; I+D+I en Tecnologías de la Información y las Comunicaciones
    The development of big data & analytics projects with the participation of several corporate divisions and research groups within and among organizations is a non-trivial problem and requires well-defined roles and processes. Since there is no accepted standard for the implementation of big data & analytics projects, project managers have to either adapt an existing data mining process methodology or create a new one. This work presents a use case for a big data & analytics project for the banking sector. The authors found out that an adaptation of ASUM-DM, a refined CRISP-DM, with the addition of big data analysis, application prototyping, and prototype evaluation, plus a strong project management work with an emphasis in communications proved the best solution to develop a cross-disciplinary, multi-organization, geographically-distributed big data & analytics project. © 2018, Springer Nature Switzerland AG.
  • No hay miniatura disponible
    Publicación
    Towards an improved asum-dm process methodology for cross-disciplinary multi-organization geographically-distributed big data & analytics projects
    (Universidad EAFIT, 2018) Angee Agudelo, Santiago; Tabares Betancur, Marta Silvia
    In recent years, the big data & analytics projects developed in big enterprises or excellence centers have special conditions like being cross-disciplinary, having participants geographically distant one another, and the participation of several organizations. This has caused that traditional methodologies used to undertake data analytics, like CRISP-DM or other emerging methodologies, be not sufficient to perform an appropriate project management. This proposal uses Design Science Research Methodology (DSRM) to identify a problem, define the objectives for a solution, design, develop and show the usage of an ASUM-DM based big data & analytics process methodology for cross-disciplinary, multi-organization, geographically-distributed work teams. The results generated are a big data & analytics project management process methodology and a gap analysis applied on three enterprise-university use cases, showing how the proposed methodology can help address the big data characteristics of a project, and coordinate and integrate several multi-organization, geographically-distributed, cross-disciplinary work teams. This process methodology is expected to ease practitioners and researchers the implementation and management of big data & analytics projects with the participation of several cross-disciplinary work teams, and geographicallydistributed organizations.