Social Network Analysis power point presentation
- 1. RATNESH SHAH SOCIAL NETWORK ANALYSISSupervisor : Prof. Abhisek Gour Presented By: Ratnesh shah
- 2. INTRODUCTION Social network analysis is : a set of relational methods for systematically understanding and identifying connections among actors. SNA provides a set of methodologies and formulas for calculating a variety of criteria that map and measure the links between things.
- 3. INTRODUCTION SNA characterizes networked structures in terms of nodes (actors, people, or things) and ties, edges, or links(relationships) that connect them. Example of social structures visualized through sna include social media network, memes spread, information circulation, disease transmission, & business networks. SNA has emerged as a key technique in modern sociology. Example biology, communication, economics, geography, political science, development studies & computer science.
- 4. SOCIOLOGICAL THEORIES Several sociological theories developed Homophily – birds of feather flock together Six degrees of separation – Milgram's experiments (1961) Strength of weak ties (1973) Spread of epidemics / conventions / news
- 5. ONLINE SOCIAL NETWORK AND RESEARCHERS Huge data readily available Volume – petabytes of user-generated content everyday Variety – text,image,speech,video... Velocity – thousands of post / minute during major events Automated data collection rather than surveys
- 6. MULTI-DISCIPLINARY RESEARCH ON OSN Computer network & distributed systems Sociology, social psychology... Network Science, complex network theory Data mining, machine learning, information retrieval, natural language processing
- 7. SOCIOLOGICAL ISSUES Sociological theories investigated on OSNs Homophily, strength of weak ties Emergence and spread of conventions OSNs different from offline SNs in some aspects Almost Zero cost of maintaining social links Important users readily connect to many ordinary ones Geographical distance does not matter
- 8. Locality of friendship in Facebook
- 9. GRAPH MODELS OF OSN Most common representation Nodes: users, edges: Social links Undirected network: Facebook Directed network: Twitter Other Variaties Network among blogs, videos.. Bipartite network, viewer-video model of youtube
- 10. NETWORK PROPERTY OF OSN Most users have few links, few have many links Presence of numerous triangle (transitivity) Small world, e.g., six degree of separation Assortativity,homophily
- 11. Friendship network among students in a US school
- 12. LINK ANALYSIS Classification of social links Strong and Weak links(e.g. based on level of interaction) Some OSN allow positive and negative links (friends and enemies) Variation of strength of links with time
- 13. CENTRALITY (IMPORTANCE ) OF NODES How important is a node in a network ? How influential is a person in a social network ? How important is a website on the web? Many proposed centrality metrics Degree centrality Closeness centrality Betweenness centrality Eigenvector centrality
- 14. UTILIZING INFORMATION CONTENT IN OSN Recommendations and search Information diffusion Spam detection /trust Authority identification Identifying news on recent events
- 16. Information viral on facebook (Information Spread )
- 17. SOCIAL NETWORK STUDY Analyze water governance : Mkindo catchment, Tanzania Analyze forest community: Biosphere reserve in chiapas,Maxico Understanding household connectivity : Village of Habu, Botswana Network governance to climate change: Swiss Gotthard region
- 18. SOCIAL NETWORK SOFTWARE UCINET The standard network analysis program, runs in windows Not optimal for large network PAJEK Program for analyzing and plotting very large networks Intuitive windows interface NetDraw Also very new, but by one of the best known names in network analysis software.
- 19. THANK YOU