Too Late to Collaborate: Challenges to the Discovery of in-progress Research
0 likes130 views
The document discusses the challenges faced by researchers in discovering ongoing, unpublished research projects, as many academics are unaware of closely related work due to reliance on traditional publication-based systems. A study involving 16 computer science researchers identified the strategies used to uncover such in-progress work, revealing a trade-off between seeking collaborators and maintaining confidentiality of their ideas. The authors propose the need for a new automated system that prioritizes real-time recommendations for ongoing research while ensuring user privacy.
![unpublished
‘dark’ literature
Problem & Research Motivation
• Existing systems are unable to support researchers in the discovery of
ongoing research efforts – i.e., work that has not yet been published.
‘gray literature’published literature
• journal articles
• conference
proceedings
• books
• preprints
• technical reports
• public talks
• public data sets
• blogs, forums, etc.
• research ideas
• manuscript drafts
• manuscripts in revision or
under review
• planned research projects
• research grants in writing
or in review
• ongoing data collection
• ongoing data analysis &
data sets
6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 6
[1] Classification scheme inspired by: Rebecca Raworth
[1]](https://image.slidesharecdn.com/2019breitingertoolatetocollaboratejcdl19-190604224954/85/Too-Late-to-Collaborate-Challenges-to-the-Discovery-of-in-progress-Research-6-320.jpg)
![Findings: Discovery-Confidentiality Trade-off
6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 12
“I think, a lot of researchers do not actually tell other people what they are currently
working on, because it's kind of a trade-off: You want to search for collaborators, but
on the other hand, you don't want to show your ideas.” (P10)
“…not everyone wants to share [ongoing research]. At least, they don't want to share it
until it's published. And then it might be too late.” (P6).
à Fear of research ideas being appropriated or plagiarized before publication
à Factors influencing trade-off: trust and timeliness
researchers wished to keep
details about their own
research confidential
researchers expressed a need
to discover ongoing research
in their field
…if you have something new, you don’t want to share too much information, because
somebody might try to copy the idea or steal the idea.” (P11)
.. especially, if it's a person that you don't know, then, of course, you don't want to tell
them like the whole thing. (P3)](https://image.slidesharecdn.com/2019breitingertoolatetocollaboratejcdl19-190604224954/85/Too-Late-to-Collaborate-Challenges-to-the-Discovery-of-in-progress-Research-12-320.jpg)
Too Late to Collaborate: Challenges to the Discovery of in-progress Research
- 1. Too Late to Collaborate: Challenges to the Discovery of in-progress Research Corinna Breitinger, Bela Gipp, Patrick Wortner, Harald Reiterer University of Konstanz & University of Wuppertal
- 2. Our Group at University of Konstanz & Wuppertal 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 2
- 3. Agenda 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 3 • Problem & Research Motivation • Study Design • Findings • Implications & Future Work
- 4. Problem 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 4 • Many academics remain unaware of other researchers who are currently working on highly related topics or research questions • Several years can pass between framing a research idea and publishing the final research paper! TOO LATE!
- 5. Problem & Research Motivation 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 5 Academic Social Networking Sites (ASNS) DLs Status quo • To identify related work, researchers typically use academic search or recommender systems embedded in DLs à only consider already published manuscripts à Researchers can choose to share details of in-progress work using ASNS - but they risk their ideas or projects being stolen à manual task; no automated or standardized procedure
- 6. unpublished ‘dark’ literature Problem & Research Motivation • Existing systems are unable to support researchers in the discovery of ongoing research efforts – i.e., work that has not yet been published. ‘gray literature’published literature • journal articles • conference proceedings • books • preprints • technical reports • public talks • public data sets • blogs, forums, etc. • research ideas • manuscript drafts • manuscripts in revision or under review • planned research projects • research grants in writing or in review • ongoing data collection • ongoing data analysis & data sets 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 6 [1] Classification scheme inspired by: Rebecca Raworth [1]
- 7. Problem & Research Motivation 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 7 Research Questions: 1) What strategies do CS researchers currently employ to discover ongoing, i.e., yet unpublished, research projects in their respective field? 2) What challenges do CS researchers face when attempting to identify ongoing research projects in their field?
- 8. Study Design Questionnaire & semi-structured interview • 16 computer science researchers (Ph.D., postdoc, Prof.) • Four domains • Human-Computer Interaction (HCI) • Information Science (Info. Sci.) • Data Analysis and Visualization (Dat. Ana.) • Software Engineering (Softw. Eng.) • Six European universities • Trinity College Dublin, TU-Berlin, University of Konstanz, TUM, Uni Münster, University of St Andrews 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 8
- 9. Study Design 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 9
- 10. Findings: Strategies used 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 10 • Researchers described individual sets of strategies to address the task of identifying in-progress research in their field: a) attending conferences (15) b) talking to colleagues (14) c) browsing researcher’s online profiles d) relying on senior researchers to inform them (5) e) attending talks (4) à face-to-face encounters (a; b; e) à personal connections (b; d) à by chance
- 11. Findings 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 11 • The ability to identify ongoing research was deemed an important task by researchers to: à avoid duplicate research à gain inspiration for their own work à identify potential collaborators “You don't want to replicate something that's already existing.” (P9) “ … someone else is working on the same thing, and then they publish it before me. Then my work is worthless.” (P1) “You want to discover ideas and develop them further. So, it's important to know what’s out there (P7) it lets you “see in which direction my research field is moving’’(P6) “The more information you have about other people, what they are actually working on, the easier it is for you to initiate collaboration. (P9) “I suppose if I could be matched with researchers doing the same thing as me that would save a lot of time. (P4) Conflict: desire to keep details about their own research confidential
- 12. Findings: Discovery-Confidentiality Trade-off 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 12 “I think, a lot of researchers do not actually tell other people what they are currently working on, because it's kind of a trade-off: You want to search for collaborators, but on the other hand, you don't want to show your ideas.” (P10) “…not everyone wants to share [ongoing research]. At least, they don't want to share it until it's published. And then it might be too late.” (P6). à Fear of research ideas being appropriated or plagiarized before publication à Factors influencing trade-off: trust and timeliness researchers wished to keep details about their own research confidential researchers expressed a need to discover ongoing research in their field …if you have something new, you don’t want to share too much information, because somebody might try to copy the idea or steal the idea.” (P11) .. especially, if it's a person that you don't know, then, of course, you don't want to tell them like the whole thing. (P3)
- 13. Implications & Future Work 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 13 • Rethink how ongoing research efforts and potential scientific collaborators can be recommended – in real time – by shifting the focus from completed research to ongoing research activity • An ‘ideal’ automated system supporting this information seeking task should: • provide timely recommendations when they matter most to researchers • ensure full confidentiality and secure storage of user’s in-progress research data to satisfy the trust requirement (cryptographic hashing) • allow the user to openly exchange more details - if they wish Currently, no system exists that considers in-progress research to retrieve related research work in a privacy-preserving manner to recommend potentially valuable scientific collaborators.
- 14. Future Work 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 14 à adapt privacy-preserving feature matching methods to researcher’s work in-progress (WIP) research artefacts à generate real-time recommendations of similar in-progress research and potential collaborators
- 15. Thank You 6/3/19 Corinna Breitinger - @BreitingerC - Challenges to the Discovery of in-progress Research 15 Contact: [email protected] Twitter: @BreitingerC Web: https://bit.ly/31bUdfN Funding provided by: