AuMixDw: Towards an automated hybrid approach for building XML data warehouses

Ouaret, Zoubir; Boukraâ, Doulkifli; Boussaïd, Omar; Chalal, Rachid

doi:10.1016/j.datak.2019.01.004

Cited by 4 publications

(1 citation statement)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…University Management requires a tool to produce information from the information records generated [4] expected to support the top-level management decision-making process. Big data can be implemented with a data warehouse to support the decision-making process [5] Information system data at higher educations each semester becomes very large data, data warehouse (dw) one of the containers for collecting a number of data in decision-making policies, which functions to store, analyze and visualize data effectively, extract transform load (ETL) processes ) [6] which functions like data collection from several sources, cannot be updated in the data warehouse optimally, so ETL must be distributed in the data warehouse for an approach to data archive reporting and visualization [7]. K-medoids or Partitioning Around Method (PAM) is a non-hierarchical grouping method in this clustering method, which is a partition to group n sets of objects into a number of k clusters, K-medoids is applied to accurately identify candidate fire areas in the Fire Detection research for surveillance applications.…”

Section: Introductionmentioning

confidence: 99%

The Implementations of K-medoids Clustering for Higher Education Accreditation by Evaluation of Davies Bouldin Index Clustering

Ghufron

Surarso

Gernowo

2020

kursor

View full text Add to dashboard Cite

The need for data analysis in tertiary education every semester is needed, this is due to the increasingly large and uncontrolled data, on the other hand generally higher education does not yet have a data warehouse and big data analysis to maintain data quality at tertiary institutions is not easy, especially to estimate the results of university accreditation high, because the data continues to grow and is not controlled, the purpose of this study is to apply k-medoids clustering by applying the calculation of the weighting matrix of higher education accreditation with the data of the last 3 years namely length of study, average GPA, student and lecturer ratio and the number of lecturers according to the study program, so that it can predict accurate cluster results, the results of this study indicate that k-medoid clustering produces good cluster data results with an evaluation value of the Bouldin index davies cluster index of 0.407029478 and is said to be a good cluster result.

show abstract