[Sigmetrics] Topic modelling in arts and humanities research
kschudamani at rediffmail.com
Sun Nov 26 02:39:56 EST 2017
I have a paper on" subject collaboration: a case study of superconductivity collaboration in Collier workshop held Nistads, new Delhi in 2001
Sent from my Samsung Galaxy smartphone.
-------- Original message --------From: Loet Leydesdorff <loet at leydesdorff.net> Date: 11/26/17 12:21 (GMT+05:30) To: Jonathan Adams <j.adams at digital-science.com>, "SIGMETRICS (sigmetrics at mail.asis.org)" <SIGMETRICS at mail.asis.org>, Tobias Hecking <hecking at collide.info> Subject: Re: [Sigmetrics] Topic modelling in arts and humanities research
I read the report with interest. We previously had some validity issues with topic modeling (Leydesdorff & Nerghes, 2017), which we discussed before. Have these issues been solved by moving from LDA to Non-negative Matrix Factorization as a method?
You claim (p. 15) that the results are "meaningful," but elsewhere you formulate more cauteously that "topic modeling is a promising approach that can capture trends in research production." "Can" or "may"?
Are the data (and perhaps some results) available online? Tobias Hecking (Duisburg) and I are trying to upscale the previous study and would love to use them as an example.
Leydesdorff, L., & Nerghes, A. (2017). Co-word maps and topic modeling: A comparison using small and medium-sized corpora (N < 1,000). [Article].
Journal of the Association for Information Science and Technology, 68
(4), 1024-1035. doi: 10.1002/asi.23740
University of Amsterdam
Amsterdam School of Communication Research (ASCoR)
loet at leydesdorff.net ; http://www.leydesdorff.net/
Associate Faculty, SPRU, University of Sussex;
Guest Professor Zhejiang Univ., Hangzhou; Visiting Professor, ISTIC, Beijing;
Visiting Fellow, Birkbeck,
University of London;
------ Original Message ------
From: "Jonathan Adams" <j.adams at digital-science.com>
To: "SIGMETRICS (sigmetrics at mail.asis.org)" <SIGMETRICS at mail.asis.org>
Sent: 11/24/2017 11:56:35 AM
Subject: [Sigmetrics] Topic modelling in arts and humanities research
Digital Science’s latest report analyses arts and humanities research through topic modelling of grant applications, unveiling an innovative and richer understanding of the research activity in disciplines that have received much less attention than STEM research.Today, working with the Arts and Humanities Research Council (AHRC), we have released our latest Digital Research Report on: “Topic Modelling of Research in the Arts and Humanities.” The report is an analysis of AHRC grant applications and details new topic modelling applications for analysing research funding data. This is the first time these methods have been applied to a national funding dataset of research grants and one of the first comprehensive analytical views across research in the arts and humanities.--
Dr Jonathan Adams
Chief Scientist, Digital Science
Visiting Professor, King's College Londonhttps://www.kcl.ac.uk/sspp/policy-institute/Visiting-professors-fellows/adams.aspx
M/ +44 7964 908449
E/ j.adams at digital-science.com
Custom reporting and analysis to help you make better decisions faster
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the SIGMETRICS