Breaking Out of the Ivory Tower: A Large-scale Analysis of Patent Citations to HCI Research - hcitransfer-chi-2023.pdf

Tags

Top Highlights

• HCI research has been cited extensively by patents — overall 20% of papers from CHI, CSCW, UIST and UbiComp, and 13.4% of SIGCHI sponsored venues, are patent-cited, including a surprising 60-80% of UIST papers over a twenty year period, higher than 1.5% of science overall and 7.7% of biomedicine;

• The patent-paper time lag is long (on average 10.5 years)

• Within HCI research, there is substantial heterogeneity in patent citations across topics, for example, interaction and input techniques research are especially likely to be referenced by patents while theory, social and experience design research are not

• not all patents will turn into products or practices, so they may not be actual “industry impact”

• the time lag between a patent and its paper citations is long (10.5 years) and getting longer, suggesting that HCI research and practice may not be efficiently connected

• One line of the literature suggests high barriers: that HCI research has remained distant from industry impact, and that “HCI researchers and HCI practitioners work in relatively separate spheres of influence” [22 ]. This line of work also argues there is a considerable research-practice gap, one that is “real and frustrating” [ 60 ] and likely the result of a long list of barriers [ 18 , 75 ].

• another line of literature argues that the field achieves considerable success, that “HCI is at the vanguard of innovation and has repeatedly influenced industry” [ 32 ] and that “there is no question that research in the area of user interface software tools has had an enormous impact on the current practice of software development” [57

• “long nose” theory on HCI innovation [12, 32].

• Patents are more likely to refer to systems- oriented and highly-cited research in HCI.

• design processes (e.g., usability testing, heuristic evaluation), design patterns, and open source software (e.g., d3, Vega Lite) also have significant industry impact that is not reflected though patents

• In fact, the ubiquity of long time delay between research and practice, and thus lack of immediate impact on the industry after the publication of a research paper, could be one underlying reason why many papers on HCI translation argue that HCI lacks practical impact [18 , 22 , 63 ]

• show that the HCI community often leaves an idea behind by the time industry gets interested, as a paper’s peak citation lag is generally shorter than the paper’s first patent citation lag

• this could also be a sign that prac- titioners are not fully aware of some application-oriented advances in academia, and that information diffusion between academia and practice is inefficient [12]

• Our work thus echoes calls for a more inclusive and translation- friendly environment [ 9, 15, 18 , 19]:

• estab- lish more communication and collaboration channels to engage people, e.g. SIGGRAPH-style Emerging Tech festivals where aca- demic researchers show their published HCI work

• volve more HCI materials in Computer Science curriculum at universities to get ‘future practitioners’ more familiar with HCI research ideas,

• For instance, in the life sciences [24 ], women faculty members patent at about 40% of the rate of men

• Encourage self-driven technology transfer.

• Encourage communications and collaborations across academia and industry.

• Recognizing translational work in HCI.