Social Network Analysis

This research paper got accepted to EDM2017

Project Description


Pressible is a school blogging and content management system developed by EdLab at Teachers College Columbia University, with over three thousand community users and more than five hundred blogs. In this paper, social network analysis and natural language processing with Latent Dirichlet Allocation topic model approaches were utilized to gain insights into Pressible, to explore four developmental stages of a college-wide social network and their associations with blog content creation. The results showed that professors who developed courses through Pressible became the most influential persons in the learning network on this blogging system. Students joined the online discussions by engaging with the course blogs to become influencers in the learning network. Students extended the online discussion topics beyond the scope of course topic set by professors. The topic cooccurrence frequencies of individuals increase as the whole learning network had more active members and connections during a rapid growth stage.

EdLab adapted the WordPress Content management systems (CMS) framework and developed Pressible for the Teachers College (TC) community in 2008. It was designed for fast content delivery, minimization of users’ time spent managing technology, and developing connections between users (Zhou, 2013). From the perspective of social constructivist theory, people communicate, contribute and acquire knowledge through social engagement and discussion of topics (Vygotsky, 1978). People also gain knowledge online via connecting information from various perspectives (Siemens, 2004). Massive Open Online Courses (MOOCs) provide more opportunities for people to study for personal intellectual growth (Kizilcec et al., 2017). Social factors from online discussion forums (Rose, et al., 2014) and engaging in higher order thinking behaviors enhanced learning in MOOCs (Wang, et al., 2016). Higher Education utilizes academic blogging to facilitate social networking, self-directed learning, and collaboration. Simulation studies on the blogosphere indicate that improved management facilities on course blogs positively affect the density and connectedness in learning networks (Wild & Sigurðarson, 2011). This study utilized social network analysis (SNA) to investigate human-human interaction and the development of social connections on this blogging platform. Next, Latent Dirichlet Allocation (LDA) topic model method was applied to understand human-information interaction during different developmental stages of Pressible. This study provides an exploratory examination of four developmental stages of an online learning community in a school blogging system.

Columbia University

Teachers College


Empty section. Edit page to add content here.

Related Projects


We're not around right now. But you can send us an email and we'll get back to you, asap.


©2018 Xiaoting Kuang

Log in with your credentials

Forgot your details?