Professor Mike Thelwall
BSc (Lancaster), PhD (Lancaster)
Information School
Professor of Data Science


Full contact details
Information School
Room C225
The Wave
2 Whitham Road
葫芦影业
S10 2AH
- Profile
-
I research scientometrics, metascience, and social media from a social science perspective. I am currently leading an ESRC-funded international metascience project assessing the value of large language models like ChatGPT for research evaluation and participating in an international project studying scientific retractions and misinformation in the media funded by the Calouste Gulbenkian Foundation. I primarily apply quantitative methods and artificial intelligence to social science issues, always with a reflexive perspective.
I previously worked at the University of Wolverhampton in 1989-2023 where I taught mathematics and statistics and researched educational technology before switching (by mistake, it鈥檚 an embarrassing story) to library and information science with a focus on web indicators for research evaluation (webometrics). I founded the Statistical Cybermetrics and Research Evaluation Group in 2000 to research bibliometrics and altmetrics/webometrics, which I led until moving to 葫芦影业. I have supervised 22 PhD students to completion in bibliometrics and sentiment analysis. My work has been cited 56,000 times and in 2015 I received the de Solla Price Medal for scientometrics.
I have collaborated on many international multidisciplinary research projects and have worked on external contracts applying innovative bibliometrics for various external organisations, including the United Nations Development Programme, the United Nations Food and Agriculture Organization, the Belgian government, Nesta (UK), UK Research and Innovation (UKRI), ESRC, Gulbenkian, and Jisc. I was part of the Metric Tide group that evaluated the role of bibliometrics in the Research Excellence Framework and now sit on the UK Forum for Responsible Research Metrics. In 2022 I led a team assessing whether artificial intelligence could play a role in future research assessment in the UK.
- Research interests
-
My core disciplinary area is bibliometrics, also known as scientometrics, using primarily quantitative methods to investigate research processes or impacts. Although researching many aspects of this field, I have partly specialised in alternative indicators for research evaluation, known as altmetrics. In the past I have investigated the use of traditional Artificial Intelligence methods for research evaluation and now I am focusing on Large Language Models (LLMs) like ChatGPT, and Google Gemini.
I have also researched sentiment analysis in the past, developing the widely used software SentiStrength, which was used by Yahoo! and other companies as well as some digital artists. SentiStrength has been used in several high-profile digital art light installations including on the London Eye during the London Olympics, and on the Empire State Building during the Super Bowl.
An important parallel and ongoing aspect of my research is developing social science research methods and applying them to a wide range of social science and humanities fields. My research is often interdisciplinary, and I collaborate with scholars in diverse fields from complexity science to Victorian studies. Themes in my research include web-based data collection, methods development and evaluation, gender analysis, research evaluation, and research on research. I see my core strengths as combining programming skills with quantitative-led mixed methods and a curiosity about current research topics and social development.
Key research outputs include software SentiStrength (sentiment analysis), Mozdeh (social media analysis), and Webometric Analysis (scientometric and altmetric data collection and analysis) and numerous specific findings and inventions, such as the Mean Normalised Log-transformed Citation Score (MNLCS) for fair and precise estimates of average citation impact. From a methods perspective, I am particularly proud of, 鈥淚鈥檓 nervous about sharing this secret with you: YouTube influencers generate strong parasocial interactions by discussing personal issues鈥 and 鈥淧redicting article quality scores with machine learning: The UK Research Excellence Framework鈥.
I would be happy to supervise PhDs related to bibliometrics or research evaluation, especially with an LLM component, as well as social media analysis topics with an emphasis on methods or large-scale data. I would also be happy to supervise broader data science projects with information science goals.
Mike's software and data is available at the following locations:
SentiStrength: and
Mozdeh: and
Webometric Analyst: and
SocSciBot: and
AI for research evaluation:
Research data:
- Publications
-
Books
Journal articles
Chapters
Conference proceedings papers
Preprints
- Teaching activities
-
INF112 - Data Modelling and Storage
INF6024 - Researching Social Media
INF6050 - Database Design
- Professional activities and memberships
-
- Committee member: UK Forum for Responsible Research Metrics (UKRI) 2017-
- Docent, Department of Information Studies, 脜bo Akademi University, Finland.
- Senior associate editor of Journal of the American Society for Information Science & Technology.
- Member of the editorial boards of:
- International Journal of Social Research Methodology (2022-)
- Profesional de la Informaci贸n (2022-)
- Data Science and Informetrics (2020-)
- Quantitative Science Studies (2019-)
- Journal of Data Science (2015-)
- Scientometrics (2007-)
- Journal of Information Science (2006-)