Please use this identifier to cite or link to this item:
|Title:||MODELING SOCIAL EFFECTS IN A CROWD-FUNDING COMMUNITY: A CASE OF PRO-SOCIAL LENDING BEHAVIOR PREDICTIONS ON KIVA.ORG||Authors:||PENG JINYUE||Keywords:||Crowd Project, Crowd-funding, Online Community, Individual Level Social Effects, Kiva||Issue Date:||19-Aug-2016||Citation:||PENG JINYUE (2016-08-19). MODELING SOCIAL EFFECTS IN A CROWD-FUNDING COMMUNITY: A CASE OF PRO-SOCIAL LENDING BEHAVIOR PREDICTIONS ON KIVA.ORG. ScholarBank@NUS Repository.||Abstract:||Currently, the idea of building online communities has spread rapidly from conventional businesses to multitudinous burgeoning online crowd projects, such as those in the crowd-funding and charity sectors. With the ever-escalating explosion of the business size of crowd projects, the evaluated value of these online communities has simultaneously escalated. The social effects in online communities is deemed an important source for user-behavior analysis and prediction in crowd projects. There is a plethora of studies examining the overall social effects in the online crowd-funding community, however, the quantification of the individual-level social effects has been investigated by very few researchers. In this thesis, based on the dataset of Kiva.org, one of the countless websites that pioneered pro-social crowd-funding, we analyze the structure of the online community and propose a feasible similarity-based prediction framework to model the individual level social influence. The experimental results indicate that the framework is useful in improving group members’ pro-social behavior prediction in Kiva. The modeling of social effects minimizes the gap on individual-level social effects analysis lacking in crowd-funding literature and could be widely adapted for user behavioral analysis in online community and online crowd projects.||URI:||http://scholarbank.nus.edu.sg/handle/10635/132150|
|Appears in Collections:||Master's Theses (Open)|
Show full item record
Files in This Item:
|PengJ.pdf||15.61 MB||Adobe PDF|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.