Please use this identifier to cite or link to this item:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/63948
Title: | 設計一考量多樣性模型以衡量使用者在社群網站上的影響力 A diversity-based model for identifying influential users in social networks |
Authors: | Chung-Ting Lin 林俊廷 |
Advisor: | 鄭卜壬(Pu-Jen Cheng) |
Keyword: | 社群網站,影響力,社群多樣性,有影響力的人,人造社群資料, social network,influence, social diversity,influence people,synthetic graph, |
Publication Year : | 2012 |
Degree: | 碩士 |
Abstract: | 社群網路(Social Network)近幾年在網路世界越來越發達,漸漸的佔據人們生活的一塊,現在多數人上網都將時間花在社群網站上面,而社群網站也成為現在人不可或缺的一部分。因為社群網站的特徵,訊息在社群網路上的傳遞速度非常快速,往往一則熱門訊息,在短短的幾個小時內,就可以透過社群網路的擴散讓各地的人可以迅速的得知消息。因此,若是能夠在社群網路上找到最有影響力的人,進而將訊息傳遞給有影響力的人,便能夠透過有影響力的人散布,而將訊息快速傳遞到很多使用者身上。
本篇主要的目的在使用社群多樣性來計算影響力,並且也提出人造社群資料來測試影響力計算。貢獻如下,1.提出先前使用PageRank演算法計算影響力時,可能會產生的spam的狀況。2.使用人造的社群檔案進行影響力的測試及評估。3.提出了數個加入社群多樣性的方法,以評估使用者的影響力。我們設計了偵測社群多樣性的方法,為計算影響力的演算法中第一個考慮到社群多樣性的概念。透過找尋有影響力的人,只要把訊息經由這些有影響力的人做為種子,那麼就可以達到快速散布訊息的目的了。 Social Network developed very fast in the online world in recent years. It has been a part of people’s life. More and more people spend a lot of time on Social Network. The hot messages spread everywhere and very fast on Social Network in few hours. Thus, if we can find the most influence people on Social Network, we can transfer the messages to the influence people on Social Network. Then the messages will spread out to many users. The purpose of this thesis is using social diversity to estimate influence. We propose a new way to evaluate our algorithm by using synthetic graph. Our contribution: 1. Point out the problems of finding influence people. 2. Use synthetic graph to evaluate influence rank. 3. Propose several methods to estimate influence with social diversity. We are the pioneer who estimate influence with social diversity. For finding the influence people , we can spread message through these people , and then the message will spread very fast and far away. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/63948 |
Fulltext Rights: | 有償授權 |
Appears in Collections: | 資訊網路與多媒體研究所 |
Files in This Item:
File | Size | Format | |
---|---|---|---|
ntu-101-1.pdf Restricted Access | 1.25 MB | Adobe PDF |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.