Chapter VI

Conclusion

6.1 ConclusionConclusion of this project are :

1. K-Means algorithm can classify the articles according to the similarity

degree well.

2. Clusters has been made can ease the readers to find articles with the

biggest similarity degree.

6.2 Further ResearchK-means clustering algorithm is an algorithm that is very good in the

grouping of the article with the greatest degree of similarity. The articles are

grouped by content owned word. Then the validity of the content of the word in

the article would greatly support the accuracy of the program. The addition of the

appropriate base word stemming process can be done on purpose to the removal

of common words.

K-Means algorithm result can be visualized by a chart. The form of chart is

scatter chart. If this algorithm has two centroid or two clusters only, it can be made

as linier regression chart. If the clusters is more than 2, the result can be visualized

throug multidimensional scalling.

36

APPROVAL AND RATIFICATION PAGESTATEMENT OF ORIGINALITYFOREWORDABSTRACTChapter IIntroduction1.1 Background1.2 Scope1.3 Objective

Chapter IILiterature Study2.1 Data Mining2.2 Clustering Algorithm2.3 Algoritma K-Means2.4 Example2.5 Data Structure2.5.1 Two Dimensional Array

Chapter IIIPlanning3.1 Research Metodology3.2 Project Management

Chapter IVAnalysis and Design4.1 Analysis4.1.1 Use Case Diagram4.1.2 Flow Chart

4.2 Design4.2.1 Class Diagram

Chapter VImplementation and Testing5.1 Implementation5.2. Testing

Chapter VIConclusion6.1 Conclusion6.2 Further Research

References