On 14th June, I met Mark E. Zulauf who is the Researcher Information Systems Coordinator of the University of Illinois at Urbana-Champagin. We met at Rm.452 of the main library on 14 June 2017 at 12:30. I sent him a list of questions in advance, for those which I wanted to have more elaborate answers and the first hand experiences. At the beginning of the interview, Mark gave me a brief history of Illinois Experts and its general routine process. Then we went through my interview questions in a Q&A style.
Before I met him, I looked into those three web pages to gather information and organize my questions.
- Illinois Experts BETA – https://experts.illinois.edu/
- Help & FAQs – http://publish.illinois.edu/researchconnections/
To introduce it briefly, Illinois Experts is the research profile database of UIUC. Two primary goals of Illinois Experts are “to facilitate discovery of potential research collaborators; and to serve as a central showcase for Illinois research, aiding in discovery.” It includes more than 1,700 STEM and social science faculty profiles and automatically collect their publication data from SCOPUS on a weekly basis. There are options to import data from other databases or add manually and the library keeps trying to augment the data with monographs and book chapters especially for the humanities units and faculty members. It doesn’t track the non-tenure line researchers’ records but will eventually include them in near future.
It uses Elsevier’s Pure which also provides the information about their research strengths and potential collaboration partners. The former is called as the Fingerprint which is automatically generated from the abstracts by following these three steps: text mining, using various thesauri, and weighting each term. If you want to look up more detailed information about Illinois Experts and Pure, the FAQ page will be extremely helpful.
My original question list was like the below, but I skipped some of them which I found the answers until a few hours beforehand or during the interview;
Q. Could you show or explain about the whole process of EXPERTS?
- How do you collect SCOPUS data?
- What kinds of queries do you use? Are they based on affiliation names? What about the researcher’s old affiliations?
- If so, how do you handle all different names for the same research units?
- Is there any recommended or official name list made by the university or the library?
Q. I saw one of the SCOPUS Author Page and there were SCOPUS Author ID and ORCID ID.
- Did library decide not to include WOS Researcher ID? or is it just because it’s Elsevier’s Pure?
- How do you handle the data which is out of SCOPUS coverage?
- Has there been any specific request for WOS based impact data? If so, how do you answer them?
- SCOPUS says its matching process (articles to author) is based on algorithm. Could you give me more detailed information about it?
Q. Is there enough room for customization in Pure?
- SCOPUS says the author or his/her organization can report to SCOPUS if the author is found in two or more document groups, and EXPERTS also says they can email you. How do you modify them? Do you need to make any modification via SCOPUS? or can you do it by yourself?
- What about the new entry items not on SCOPUS? (e.g. the humanities researchers’ works from alternative sources)
- Could you tell me the alternative sources (if there are more than OCLC WorldCat) and how you collect them?
Q. After collecting data, how do you clean them?
- Is there any settled process for people to clean the data? If it’s changeable, could you give me some case examples?
- How do you categorize the works? Is there any criteria? or is it set by SCOPUS
- How does Pure select Fingerprints? By which rules does it apply to weight certain keywords?
Q. How much time and resources have you put into EXPERTS (including Illinois Research Connections)?
- How come did you decide to subscribe Pure? Was there any other options?
- What advantages and disadvantages does Pure have?
- Is the working with Elsevier going well? Such as, is it supportive and good at trouble shooting?
- Could you tell me your best/worst experiences in developing it? or any advice or comment?
Q. If there are more future plans other than that on the website, what it would be like?
The beginning of Illinois Experts, called Illinois Research Connections at that time, goes back to 2013. It was originally from the Campus Strategic Plan and the discussion had been going on until 2014, when the library decided to subscribe Elsevier’s Pure. The decision was made partially because of the library’s pre-existing relationship with Elsevier and mainly to save time. The library and the Office of the Vice Chancellor for Research did not want to spend longer time on in-house development but wanted to launch it as soon as possible.
In 2015 Fall, the system was set up and launched. At first time the access was restricted to campus but opened to world in 2016 Spring. It started with 1,700 faculty profiles and now the library just have completed building the profiles of 500 non-tenure line researchers (emeritus faculty, clinical faculty and academic professions.) They are not yet opened to public with other 320± profiles in humanities and arts units.
I will summarize what we talked during the later part of the interview, mainly which are not described on the web pages listed up above. Though there were some unknown things and internal information (e.g. the algorithm that Pure uses to weights research terms and the subscription fee), the time with Mark gave me a better understanding on how Pure works and how his team manages it.
- Before the system starts its actual weekly updating, the university gives the human resource information to Pure Profile Refinement Team and the team review it manually to increase data accuracy and reliability. During this manual reviewing process, some non-public information, such as the information of the researcher’s past affiliations and more common names, are being included.
- For research units, there is a “Campus Code Book” which is established in Pure. Each unit has a code for the unit’s official name and also its researchers’ names to connect them but it doesn’t include the name variations. Researchers can add their different names on Pure but it’s not automatically reflected to the SCOPUS data. To do that the researcher him/herself or the librarian have to contact SCOPUS.
- Besides the databases mentioned on the web pages (Web of Science, Mendeley, WorldCat, CrossRef, and Google Scholar) now it became available for researchers to import the data from EBSCOhost databases manually. Still, its automatic import module only works with Mendeley and Web of Science. Pure has more import modules, such as CV module, Report module, several funding modules too, but the library doesn’t subscribe them.
- There is no data cleaning process for the SCOPUS data because of the shortage of staff. Actually he is the only one full time employee for this project. He had worked with one graduate student who worked 13 hours per week, and last semester he had three students (equivalent to 1 FTE) helped him to collect publications and arts data in humanities and arts.
- Working with PURE Team is pretty okay. It responds quickly and sometime they accept some suggestions to enhance the system (e.g. UIUC’s suggestions on NewsFlow feed feature, which was released last fall.) There are more than 50 institutes using Pure in North American User Group and they have annual meetings too.
- The best thing in working on Illinois Experts is a lot of interests is there. Communicating people at department units are fascinated and use its data to promote their units. Illinois Experts does promote scholarships and research activities and facilitate the connection with researchers. It also keeps more research funding and collaborations among the researchers. On the other hand, the biggest challenge is the lack of staff. With more human resources, the data quality control would become more feasible.
To me, it looks possible to import data from other databases or other reference management software because it supports manual RIS file import for Google Scholar. The data exported from Google Scholar is barely enough to create citations and a bibliography so it would be worth to test whether or not the researchers can import the RIS files exported from other academic databases.
I have wanted to talk with the people doing the same job at the other universities because I am working on the similar project at my home library. My library set the plan to build a faculty profile database this year and a tailored research information recommendation system the next year, so it was a great opportunity for me to interview with Mark. I do appreciate Mark and Myeong-Ja Han (who introduced me to him) for their time and help, and I hope I could provide more in-depth observations and information to my team and the wider librarian community in South Korea.