Please use this identifier to cite or link to this item:
|Title: ||ä»¥ç¸é°ç¾¤é«å°æ¾éçç¤¾ç¾¤æ¼è¤éç¶²è·¯ä¹ç ç©¶|
A Study on Developing an Overlapping Community Detection Method for Large Directed Social Networks
Social Network Analysis;Community Identification;Overlapping Community;1.5-club
|Issue Date: ||2011-05-24 15:11:59 (UTC+8)|
|Abstract: ||èªWeb 2.0å´èµ·ï¼ç¶²ç«å»ºç½®éå§ä»¥ãä½¿ç¨è
¶äººééä¿ç¶²è·¯å å¼ç¼ç¾¤èææï¼é²èå½¢æèæ¬ç¤¾ç¾¤ãè¿å¹´ä¾ééç¤¾ç¾¤ç¶²è·¯åæ(Social Network Analysis; SNA)å¯ä»¥ææå°æ¾åºé±å«å¨å¤§åäººéç¶²è·¯å
ç±ååè¿ä¼¼æ³ CPM(Clique Percolation Method)ä¾éæ¥åä½µæ»¿è¶³1.5-clubæ¢ä»¶çç¯é»ï¼å¨æ¾åºåæ¥çåç¤¾ç¾¤çµæ§å¾ï¼åçµåç¸é°ä¸è½å¢å åç¤¾ç¾¤å¯åº¦ä¹å¿ èª ç¯é»å°åç¤¾ç¾¤ä¸ä»¥æååç¤¾ç¾¤ä¹å¯åº¦ï¼æå¾ä»¥Modularityæ¹å¼ä¾é©èåç¾¤å¾ä¹åç¤¾ç¾¤çåè³ªã|
Since the genesis of Web 2.0, websites increasingly adopt a âuser-centeredâ structure. Growing interpersonal networking leads to the clustering effect and then gives rise to virtual communities. In recent years, Social Network Analysis (SNA) has been found to be capable of effectively identifying virtual communities embedded in large interpersonal networks. Though SNA results can shed light on a lot of Internet phenomena, most research divided up the Internet to find non-overlapping groups. In reality, however, it is inevitable for there to be overlaps between groups. In view of these overlaps, this paper adopts the regional approximation method proposed by the CPM (Clique Percolation Method) in order to gradually merge to meet the 1.5-club node conditions. After a subset of the initial structure is identified, it is combined with adjacent sub-agencies and loyal groups to increase node density, and eventually community density. Finally, Modularity is applied to verify sub-grouping qualities.
|Appears in Collections:||[資訊工程系(所) ] 博碩士論文|
Files in This Item:
There are no files associated with this item.
All items in STUAIR are protected by copyright, with all rights reserved.