Probabilistic Partnership Index (PPI) in Social Network Analysis using Kretschmer Approach Nisa Sharafina CSF 1 , W. Maharani 2 , Adiwijaya 3 , N. Taniarza 4 School of Computing Telkom University Bandung 40257, Indonesia 1 sharafina.nisa@gmail.com , 2 wmaharani@gmail.com , 3 adiwijaya@telkomuniversity.ac.id , 4 natyataniarza@gmail.com Abstract— Nowadays, Twitter has become an effective media to communicate as the increasing number of its user. The interaction or relation formed in Twitter could be represented into a graph and calculated using centrality measurement method. Centrality measurement can be used as parameter to determine the popularity or leverage level of an actor towards other actor. The value of centrality measurement is a weighted graph. To maximalize the result, every relation in a graph will be added value from Probabilistic Partnership Index (PPI) method calculation. Furtherly, the analysis and implementation with degree centrality are executed with Kretschmer method using the value from PPI measurement. From the conducted experimental process value, we observed that PPI and Kretschmer can be used as one of the centrality method to determine the leverage level and popular actor of an environment in Twitter. Keywords— Probabilistic Partnership Index (PPI); social network; degree centrality; Kretschmer; I. Introduction Recent years, the use of social media in the internet has increased incisively as the increase number of the users. Twitter is one of most popular social media nowadays. Based on the survey held by a third party Twitter analysis company, Twopcharts.com, user of Twitter had exceeded 1 billion users (this data taken on January 19 th 2013). Within this condition, there are many relations formed by users. The relation between actors in Twitter can be used to represent its interaction. Then, this interaction can be measured and tested to perform its analysis or furtherly known as social network analysis (SNA) [6] [11]. Centrality measurement is one of the SNA measurement methods which can help to measure and determine the importance of relations between nodes in the network [3]. Centrality measurement consists of three different centrality measures, which are degree centrality, betweness centrality, and closeness centrality. But, this paper will only focus on degree centrality. Degree centrality is a method for measuring the number of ties/links a node has [10]. For more specific, degree centrality measure which will be used is degree centrality method introduced by Kretschmer. Kretschmer’s degree centrality method is new complex measure of degree centrality including weighted ties. To differentiate, the previous degree centrality method was using unweighted ties. On bibliometrics or webometrics analysis, there are four methods, known as collaborative linkage indexes, which can be used to measure the strength of nodes. The collaborative linkage indexes are Jaccard Index, Salton-Ochiai Indexes, Probabilistic Afinity Index (PAI), and Probabilistic Partnership Index (PPI). But, the objective in this paper is to analyze the degree centrality of social network Twitter using Probabilistic Partnership Index (PPI) indicator. Based on previous experiment, PPI can be used for measuring scientific lingkages between two countries, Japan and France. From conducted experiment, the result showed that PPI is useful in examining individual networks within complex exchanges [12]. Hence, using PPI as well as Kretschmer’s method can help the analysis of social network to determine the popularity level of an actor in a community. II. METHODOLOGY A. Social Network Analysis Social Network Analysis (SNA) is defined as the mapping and measuring of relationships and flows between people, groups, organizations, and other connected information by Krebs [11]. While L.C. Freeman said that SNA is the techniques focusing on uncovered patterns of people’s interaction [6]. Scott also has another definition for SNA that is a set of methods for the investigation of relational aspects of social structures [15]. Of those opinions, it can be concluded that SNA is a method that can be used to analyze the social interactions within a group of people by looking at the behavior of those people while interact each other. The behavior between individuals can be seen not only in real world but also from the social network, like Twitter. It can be determined its structures and also patterns of the interactions. B. Probabilistic Partnership Index (PPI) Measurement Probabilistic Partnership Index (PPI) measures the scientific linkages which employ Monte-Carlo method for evaluating the expected values and standard deviations of the number of nodes of all patterns of cooperation [12]. PPI is formulated as follows: 2014 2nd International Conference on Information and Communication Technology (ICoICT) 454