Blog Post

What are HASTAC Scholars 2013-2014 talking about?

What are HASTAC Scholars 2013-2014 talking about?

I've been talking to Fiona and Ruby about the multitude of interests of HASTAC Scholars this year. So in this post I tried to find the major areas of interest across the two hundred-plus scholars that joined this year. The primary data set was provided by Fiona and it includes a table of users with geographic information and the topics of interest to each user. I used the Scholars biography, project, and keywords to map the relationship between Scholars and topics. I modeled the network in a simple fashion, with scholars and topics as vertices, and connections between scholars and topics as edges.

Social network analysis uses the term nodes to refer to people and the term edges to refer to the interactions between them. The term "edges" is derived from geometry where an edge is a line segment joining two adjacent vertices in a figure. An edge is thus a link between points usually depicted as a line segment. In this graph we show that nodes (Scholars) are connected (edges) to topics and the more connections a topic has, the bigger is the node size in the graph. In short, this is a graph in which Scholars are linked through common topics

Among other things, this graph enables the identification of topic communities – that is, areas of interest with intra-group edge densities that are higher than expected. You can click on the graphs to see a larger image, or else you can access the full graph here. To put it simply, it allows us to see what are the common interests shared by Scholars 2013-2014. This might come as a surprise for some of you, but Digital Humanities and Social Media are the most connected topics across Scholars 2013-2014. These are followed by Pedagogy, Gender Studies, Games, and Postcolonialism.

It might come as a surprise because these are not the most common topics listed in the Scholar's biographies and projects. The interesting thing about this graph is that edges are derived from three different sources. The first one is the Scholars projects, the second is the Scholars short biography, and the third one is the Scholars list of interests (keywords). As you zoom in, the graph will show that each edge has a label indicating the source of the connection. These sources are named project, bio, and keyword, and colored red, blue, and green, respectively. There are far more edges coming from keywords than from biographies or projects (see pie chart below), so we had to assign different weights for each source of information.

The difference is interesting because it contrasts Scholar's current and past projects with their common areas of interest. The graph shows a large concentration of red edges at the center of the graph. These edges are drawn from Scholars projects and it is not surprising that they converge to "digital humanities" and related areas. After all, these are the topics that bring Scholars together despite the diversity of interests. Most of the common topics result from the projects Scholars are undertaking, with top topics (nodes with higher degree) being "new media," "digital," and "hastac." This is similar to edges based on Scholars biography (colored blue), which includes "new media," "university," "digital," and "history" as the most common topics.

The larger and more disperse edges colored green are pushed to the periphery of the graph. This indicates common topics based on keywords provided by the Scholars. There is more green in the graph because most connections come from this source of information, but it is also interesting to see that the keywords provided by the scholars is so diverse that it is proportionally more peripheral than topics clustered by projects or biographies. However, the keywords are responsible for showing the sheer diversity of topics, particularly "gaming," "literature," "social media,"  "gender studies," "pedagogy," "libraries," "rhetoric," "postcolonialism," "films," and "history." 

 

The next thing to analyze is how topics of interest to HASTAC Scholars relate to the geography of the country. I'll get back to that in future posts.

 

This material is based upon work supported by the National Science Foundation under Grant Number 1243622. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

 
50

10 comments

Thanks for sharing this Marco!   I was surprised at the prevalence of some of the topics, and reassured by others.  (It's nice to know there's other folks out there with similar passions!)  

I especially appreciated being able to distinguish between interests via keywords, projects and bios. 

I'm curious to know what software you utilized to create the visualization.  Are there any resources you would recommend for conducting social network analyses? 

Thanks!

60

Hi Amy,

I'm hoping to post a follow up to this tomorrow and that might interest you. I used igraph to generate the networks and Gephi to plot the graph (the visualization). The large plot is a Seadragon export but I'm kinda moving towards sigma.js, so the next plots might be plain HTML5. If you're interested in SNA, I think it's worth taking a look at the SAGE Handbook of Social Network Analysis (http://www.uk.sagepub.com/refbooks/Book232753) and playing with Gephi juts for fun. Here's a straightforward tutorial to plot you own Facebook social network: http://www.slideshare.net/persuasion/facebook-network-analysis-using-gephi-11390089

Hope you have some fun.

Cheers,

Marco

46

I second Amy in thanking Marco for a really visually rich map of the HASTAC scholars' interests.  I find it especially encouraging to see "gender studies" showing up with some significance. I love too how it gives us a sense of the vocabulary that we're using to articulate these different interests.  Thanks for sharing this!

43

Hi Jordan,

Gender studies is indeed a very connected topic. I'll get back to this in the next posts.

Cheers,

Marco

43

Hi Marco,

love the analysis and especially appreciate you sharing the software you are using. I look forward to learning more from your posts later this week. I do have a question -- how are thematic analysis and SNA different?  the  interactions?  so fascinating... perhaps by capturing a classroom discussion, you could visualize how students interact, the kinds of questions they astk and perhaps study the network analysis and patterns of the more successful class discussions (by measuring against learning outcomes somehow). 

thank you!

Angela

 

46

Hi Marco,

You’re most welcome and do ask me anytime you have a question about the analysis. I do most of my work using R, which is an open source platform for statistical computing. That means all the coding can be replicated by other users. The analysis of topics is not SNA per se and it’s usually part of natural language processing and machine learning (not quite this case, but still). The aim of topic modeling is to find the topics that occur in a sample of data without going thru the trouble of actually reading the entire collection. There’s no directionality between topics, but rather co-occurrence and similar frequencies based on likelihood fit. This case was particularly different because the keywords were attached to users, so it was possible to plot the relationship between topics as an affiliation network. It would be very cool to see a real-time affiliation network of students’ interaction in the classroom. It's theoretically possible, but it should be tricky to get the data.

Cheers,

Marco

55

Thanks so much for sharing these resources Marco!

47

Let me know if you have any questions, Amy.

40

Anonymous (not verified)

hi......

You’re most welcome and do request from me anytime you have got an issue concerning the analysis. I do most of my work exploitation R, that is associate degree open supply platform for applied mathematics computing. meaning all the committal to writing may be replicated by different users. The analysis of topics isn\'t SNA intrinsically and it’s typically a part of language process and machine learning (not quite this case, but still).
-----------------

[url=http://www.ezdia.com/epad/ati-radeon-hd-5670-series-graphics-card-amd/69... RADEON HD 5670 – SERIES GRAPHICS CARDS FROM AMD[/url]

46

Please remove last comment from Anonymous which is spam

57