Liu Y. (2014) Identity Crisis of Ubicomp? Nonsense!

Paper: Identity Crisis of Ubicomp? Mapping 15 Years of the Field’s Development and Paradigm Change (2014)
Author(s): Yong Liu et al.

The authors have ambitiously set themselves the task of “mapping 15 years of the field’s development and paradigm change[s]”, the field in question being ubiquitous computing. Liu et al. have done this very aptly in their work. The overarching thesis of the paper is to refute the assertion that Ubicomp research is suffering an identity crisis. They felt that Abowd’s assessment that the field of ubicomp is losing its unique identity as a result of being inherently interdisciplinary [1]. Liu et al. argue that this is a subjective and an individual (and possibly biased by other factors, which the authors do not mention) point of view. Refuting this assertion is the authors’ primary motivation. The authors support their statement by developing a supervised algorithm to extract keywords from the abstract of 1636 papers published in the 15 years (1999-2013) from the main Ubicomp conference series (UbiComp, Pervasive, HUC).

Strategic diagrams (two-dimensional plots used for co-word analysis) were used to show the two key concepts identified by the authors as useful to map the field of Ubicomp: density and centrality. Keywords contained in a paper were linked together to form a cluster (a set of closely-related keywords). Density, or internal cohesion measures the strength of the links that tie together the cluster of keywords. Centrality measures the degree of interaction of a theme with other parts of the network. Their study found a noticeable paradigm shift between the periods  of 1999-2007 and 2008-2013 not only in terms of growth but also in the increase in cohesion within the field. They point to the increasing network density and average cluster centrality of the last 15 years in their support. This they argue, wouldn’t have happened if the field was fading. A correspondence analysis of keywords and year done by the authors, similarly shows a clear trace showing that consecutive years are located close to each other, and the absence of an abrupt change between consecutive years.

The researchers were heavily involved in curating and culling keywords/clusters as necessary, relying on their judgment on individual keywords. This does open them up to the same charge of possible bias (admittedly at a smaller scale, due to their thorough analysis) that they accuse Abowd and other proponents of the identity crisis claim. Another aspect to consider, which the authors preempted in their paper, is that the papers submitted to these 3 mainstream conferences may not be representative of all research undertaken in this field, especially taking into account their low acceptance rates. Indeed, one of Abowd’s assertions in favour of the identity crisis stance is that papers published in the UbiComp conference could have appeared in one or more of other conferences or journals that do not align with ubiquitous computing.  I felt that this objection was not sufficiently addressed by the authors. All in all, their systematic undertaking is convincing and does show that a blanket dismissal of ubiquitous computing is premature and unwarranted.

The take home point for me is that despite the prevalence of ubiquitous computing, especially in fields traditionally not associated with computer science, as an independent field it is still relevant. There may be shifts in focus and there may be fields once considered core elements of ubicomp that have been “dropped” from the list of topics, but that is not a reason to throw out the baby with the bathwater.

The paper i would like to recommend as an example of ubiquitous computing technology is:
[1] G. D. Abowd, A. F. Bobick, I. a Essa, E. D. Mynatt, and W. a Rogers, “The aware home: A living laboratory for technologies for successful aging,”
Proc. AAAI-02 Work. Autom., pp. 1–7, 2002.

My reason for choosing this is that it was a long term research program utilising ubiquitous computing to create a living laboratory. This lab was used as the base for research into maintaining quality of life for an ageing population.

References
1. Abowd, Gregory D. “What next, ubicomp?: celebrating an intellectual disappearing act.” Proceedings of the 2012 ACM Conference on Ubiquitous Computing. ACM, 2012.