Implementation of MCL algorithm in clustering digital news with graph representation

Digital news will continue to grow up and evolve, it bring up the new issue for modeling digital news data that are stored in the database so that it is easier to understand and be able to take some important information thoroughly. To simplify the information processing in the database it is require a model and a specific method for clustering the news based on proximity and characteristics of the digital news. Using a graph database models and methods, MCL graph clustering algorithm (Markov Cluster Algorithm) can simplify information processing by identifying the characteristics of each vertex in the graph so that it will establish cluster vertices with a specific label. In the process of identifying a clusters of graph, the digital news documents will be stored into one vertex to be connected with other vertex in common category of news. The process of expanding and inflating matrix will be the main process in the clustering of digital news that has been transformed into a graph database models that expand aims to show a new edge and remove the old edge in common not needed in the graph. Meanwhile the process of inflating aims to strengthen the strong edge and weaken the weak edge. So that with the clustering of digital news, process of inflating matrix is very influential for execution time in MCL algorithm and process of inflating matrix influence the number of cluster will be formed.