, Volume 106, Issue 3, pp 1135–1150

A Sciento-text framework to characterize research strength of institutions at fine-grained thematic area level

  • Ashraf Uddin
  • Jaideep Bhoosreddy
  • Marisha Tiwari
  • Vivek Kumar Singh


This paper presents a Sciento-text framework to characterize and assess research performance of leading world institutions in fine-grained thematic areas. While most of the popular university research rankings rank universities either on their overall research performance or on a particular subject, we have tried to devise a system to identify strong research centres at a more fine-grained level of research themes of a subject. Computer science (CS) research output of more than 400 universities in the world is taken as the case in point to demonstrate the working of the framework. The Sciento-text framework comprises of standard scientometric and text analytics components. First of all every research paper in the data is classified into different thematic areas in a systematic manner and then standard scientometric methodology is used to identify and assess research strengths of different institutions in a particular research theme (say Artificial Intelligence for CS domain). The performance of framework components is evaluated and the complete system is deployed on the Web at url: The framework is extendable to other subject domains with little modification.


Computer science research Research competitiveness Field-based ranking Scientometrics UniversitySelectPlus 

Copyright information

© Akadémiai Kiadó, Budapest, Hungary 2016

Authors and Affiliations

  • Ashraf Uddin
    • 1
  • Jaideep Bhoosreddy
    • 2
  • Marisha Tiwari
    • 3
  • Vivek Kumar Singh
    • 4
  1. 1.Department of Computer ScienceSouth Asian UniversityNew DelhiIndia
  2. 2.Department of Computer Science and EngineeringUniversity at BuffaloBuffaloUSA
  3. 3.DST-CIMSBanaras Hindu UniversityVaranasiIndia
  4. 4.Department of Computer ScienceBanaras Hindu UniversityVaranasiIndia

Personalised recommendations