I was avoiding real work online and I stumbled across an interesting area of Google Scholar. Turns out they’ve calculated the h5-index of various publication by subfield and HCI is one of the sub-areas featured. It’s obviously not perfect, for example, CHI appears as two different conferences depending on how people cited it, but it’s still fascinating.

First, I’m especially grateful that this list has exposed me to a publication with which I wasn’t familiar but that I find quite relevant and fascinating: the journal of Computers in Human Behavior. Just a quick skim through the top 20 papers suggests that there might be a lot of good stuff that I’ve been overlooking in my searches. I do wonder why a lot of these papers haven’t appeared on my previous related work investigations (of say, video game addiction)? Am I not using the right search terms?

Second, I found it interesting that the results don’t seem to closely match another list gathered of average citations per paper. I tend to trust the Google results more as they have a more nuanced metric and the results seem more believable (at least in terms of CHI being very high on the list and the major conferences being represented). Also, the Google list only includes the last 5 years of publications rather than the whole corpus of each venue, so I think it gives a more accurate idea of the way things currently stand.

Lastly, I wanted to do a quick clustering of the top ten cited papers in CHI, CSCW, and UbiComp over the last 5 years to figure out what we’re citing the most. I picked these 3 conferences because I’ve actually been to these, so I feel a bit more comfortable doing this classification.

• CHI seems to care the most about social networking sites (3/10 papers). Otherwise, there is a lot of variety in the top cited papers, including: crowdsourcing, activity sensing to support fitness, surface computing, and end-user programming. There were else some reflection and vision papers on this list, which wasn’t the case for the other two conferences. I guess CHI especially likes talking about the big picture.
• CSCW seems to also mostly care about social networking sites (5/10 papers), especially Twitter (3 out of those 5). Disaster response, social search, crowds and wikipedia, and input devices for collaboration each have some representation on that list too. But, I’m most psyched that family videochat is on that list as well. It’s not my paper, but it’s nice to know that people care about my topic. Though CSCW has been exploring non-work domains, it is still the only conference with any top papers explicitly focusing on work environments (2/10 papers).
• Ubicomp is obsessed with sensing what people are doing and where they are doing it, with 3 papers on activity recognition, 2 on location sensing, and 1 on sensing events over power lines. But apparently, Ubicomp also cares about useful contexts for sensing with papers about sensing activities or location is order to support family awareness, sustainability, and health. This is confirming my earlier assertion that it should change its name to SensorComp.

I hope you find these lists as interesting as I did. Were you surprised by anything you saw (or didn’t see) there?