Thashmee Karunaratne

Thashmee Karunaratne


View page in English
Arbetar vid Institutionen för data- och systemvetenskap
Telefon 08-16 16 05
Besöksadress Nodhuset, Borgarfjordsgatan 12
Postadress Institutionen för data- och systemvetenskap 164 07 Kista

Om mig

I am a project manager and a researcher at the Department of  Computer and Systems Sciences (DSV), and currently conducting research in the field of Learning analytics and educational data mining. I am also within ICT for Development group. The main focus of my research is on how to use the log data in the e-learning systems as another dimension for student and course evaluations, mainly for online courses. I am also involving in investigations on how and in which ways ICT has been useful to bridge in communication and collaboration gaps. I also am the project manager of Sida-funded project Capacity building in higher education in Rwanda and EU/EP7 funded H2020-MSCA-RISE-2014  project on Transitional Diaspora Entrepreneurship as a development link between home and resident countries.  I also teach various courses and supervise students at DSV.

I have a PhD in Computer Science (Machine Learning and Data mining) 

(Thesis: Learning predictive models from graph data using pattern mining)  My PhD research work was mostly related to finding simpler and efficient methods to learn from complex and structured data.

Personal web:


I urval från Stockholms universitets publikationsdatabas
  • 2018. Thashmee Karunaratne. Electronic Journal of e-Learning 16 (2), 79-90

    The thesis component of a degree program is vital since the quality of it contributes to the quality of the whole degree. Maintaining the quality of the degree programs and handling the constantly increasing numbers of students entering higher education simultaneously is a challenge for many higher educational institutions. This paper presents a study of how ICT can be used to improve the quality and effectiveness of the thesis projects at Bachelors and Masters Levels. Further, how the blended model of supervision supports solving the issues of managing supervisor time efficiently and providing a quality guidance for thesis students are also explored. Supervisors’ perceptions of the ICT enabled thesis process are captured via interviews. Statistics about the completed theses and the user log data of the ICT system are triangulated to complement supervisor perceptions. Results revealed that the supervisors take advantage of the functions in the system to support improving the quality and the quantity of the theses, and the blended supervision model adapted in the thesis process support the supervisors to have a better collaboration with the students.

  • 2018. Thashmee Karunaratne, Ranil Peiris Colombage, Henrik Hansson. ijEDict - International Journal of Education and Development using Information and Communication Technology 14 (1), 118-140

    ICT and development implementations in developing regions depend on many factors. This paper summarises experiences of efforts made by twenty individuals when implementing small-scale ICT development projects in their organizations located in seven developing countries. The main focus of these projects was the use of ICT in educational settings. Challenges encountered and the contributing factors for implementation success of the projects are systematically investigated using interviews and follow up surveys. Results show that the typical limitations of technology and infrastructure were the key obstacles. The commitment of individual project managers in the role of “change agents” and organizational support were the strengths behind the success of the projects. Based on the outcome of this study, professional development of the change agents is a key factor for the success of projects. IT and infrastructure limitations contributed to the failure of the majority of the ICT related projects.

  • 2017. Thashmee Karunaratne, Henrik Hansson, Sten Holmberg.

    At universities students and supervisors strive for interesting and meaningful research problems to study in thesis projects. For students, finding a good topic for their thesis work is a lonely and tiresome process. A meaningful topic motivate supervisors to closely monitor and support students. The industry, state and the civil society where the interesting research problems exist on the other hand struggle to find appropriate expertise to solve their practical problems. Although connecting right stakeholders together fulfil the requirement of all the counterparts, such a collaboration is not as simple as it sounds. As Hansson (Hansson, et al., 2014) acknowledges, current efforts of connecting both the entities are mostly ad hoc and driven by private and personal networks. Many initiatives such as, Kista science city project (Kista Sceince City, 2017), Urban ICT arena (Urban ICT Arena, 2017), Samverkancheck (Collaboration) project at Stockholm University (Stockholm University, 2017) and many small scale collaboration initiatives of Ericsson and IBM with Swedish Universities can be named as few successful attempts. Silicon Valley in United States is another popular example for similar collaboration. With the use of information systems to manage and support activities, research collaborations are empowered by platforms that search and connect people for research and development purposes. Idea bank is a popular initiative for fulfilling this requirement. However, idea bank initiatives have limitations of connecting interested people to experts from a wider society. But there is a huge demand especially from developing regions, for new ideas which can be used as thesis works in higher education. A large number of students in these countries still struggle to find interesting and meaningful ideas to work with in their thesis work, while the society, state and industry are compelled for expensive problem solving approaches. Therefore a generic and globally reachable approach is required to fill in this gap of connecting required expertise with the needs. The Global idea bank approach is motivated by the exact requirement. The primary aim of this approach is to connect different counterparts in solving problems of general and practical interest. Thereby following objectives are set. 1. Developing an entry point for big and small ideas to be investigated by universities, 2. Creating supportive processes to facilitate communication among the stakeholders 3. Developing functions to create, select, sort, improve, refine, and filter ideas 4. Developing a gateway to connect the ideas to thesis support systems (eg. SciPro (Hansson & Moberg, 2011)) at universities 5. Facilitate the process with relevant information and -online templates for reporting and finance management

  • 2017. Thashmee Karunaratne, Henrik Hansson, Naghmeh Aghaee. Assessment in education

    Improving the quality of Bachelor’s and Master’s theses while at the same time increasing the number of theses without expanding the existing resources proportionately is a huge challenge faced by higher educational institutions. The aim of this study is to investigate the effect of multiple change processes on Bachelors and Masters level thesis work in a selected higher educational institution. The following research questions were studied: (1) How has the thesis quality changed? (2) How has the number of completed theses changed? and, (3) How has the ratio of completed theses per supervisor changed? The change processes were introduced into the thesis process in the Department of Computer and Systems Sciences (DSV), Stockholm University during 2008–2014. The results show that the quality and the number of completed theses have significantly increased. The multiple change processes including a purpose built ICT system named SciPro, which was introduced and improved incrementally during 2010–2014 are discussed and evaluated in relation to these results.

  • 2017. Thashmee Karunaratne, Jean Claude Byungura. IST-Africa 2017 Conference Proceedings
  • 2016. Jean Claude Byungura (et al.). IST-Africa 2016 Conference

    Integrating technology in pedagogy is a step for ICT capacity building for higher education to meet its current demands. Therefore, the integration of eLearning systems has been problematic, albeit huge investments in ICT infrastructure. This study investigates teacher adoption of a new upgraded eLearning platform being integrated at University of Rwanda. A six-constructs model related to technology adoption was used to design questionnaire and interviews. Closed and open-ended questions seeking perceptions on the UR eLearning environment were used on 87 respondents who were purposively selected. Findings indicate that although participants find the system useful, easy and trustworthy, the intention for adopting and using it is very low due managerial support and technical support. Gaps in policy synergy, incentives, basic infrastructure, managerial and technical support were among the identified bottlenecks contributing negatively to the low degree of teacher intention. The study concludes by proposing some remedies to address the above challenges.

  • 2016. Jean Claude Byungura (et al.). European Journal of Open, Distance and e-Learning 19 (2), 46-62

    With the development of technology in the 21st Century, education systems attempt to integrate technology-based tools to improve experiences in pedagogy and administration. It is becoming increasingly prominent to build human and ICT infrastructure capacities at universities from policy to implementation level. Using a critical discourse analysis, this study investigates the articulation of ICT capacity building strategies from both national and institutional ICT policies in Rwanda, focusing on the higher education. Eleven policy documents were collected and deeply analyzed to understand which claims of ICT capacity building are made. The analysis shows that strategies for building ICT capacities are evidently observed from national level policies and only in two institutional policies (KIST and NUR). Among 25 components of ICT capacity building used, the ones related to human capacity are not plainly described. Additionally, neither national nor institutional policy documents include the creation of financial schemes for students to acquire ICT tools whilst learners are key stakeholders. Although there is some translation of ICT capacity building strategies from national to some institutional policies, planning for motivation and provision of incentives to innovators is not stated in any of the institutional policies and this is a key to effective technology integration.

  • 2016. Naghmeh Aghaee (et al.). The International Review of Research in Open and Distributed Learning 17 (3), 360-383

    Many research studies have highlighted the low completion rate and slow progress in PhD education. Universities strive to improve throughput and quality in their PhD education programs. In this study, the perceived problems of PhD education are investigated from PhD students' points of view, and how an Information and Communication Technology Support System (ICTSS) may alleviate these problems. Data were collected through an online open questionnaire sent to the PhD students at the Department of (the institution's name has been removed during the double-blind review) with a 59% response rate. The results revealed a number of problems in the PhD education and highlighted how online technology can support PhD education and facilitate interaction and communication, affect the PhD students' satisfaction, and have positive impacts on PhD students' stress. A system was prototyped, in order to facilitate different types of online interaction through accessing a set of online and structured resources and specific communication channels. Although the number of informants was not large, the result of the study provided some rudimentary ideas that refer to interaction problems and how an online ICTSS may facilitate PhD education by providing distance and collaborative learning, and PhD students' self-managed communication.

  • 2015. Nam Aghaee (et al.). Proceedings of E-Learn, 237-244

    The low completion rate and slow progress in PhD education have been highlighted in many studies. However, the interaction problems and communication gaps that PhD students encounter make this attempt even more challenging. The aim of this study is to investigate the peer interaction problems and ICT based solutions from PhD students’ perspectives. The data collection method was an online questionnaire and in-depth interviews were used to follow up. The target group for the survey was the PhD students in Computer Science at Stockholm University. The total number of respondents for the survey was 53 PhD students and eleven randomly selected PhD students for the interviews. The results reflected a lack of peer interaction as an important issue in the perspective of the students. Based on this, the study showed several ICT solutions that have the potential to reduce the interaction problems and thereby improve PhD students’ collaborative learning and research quality.

  • 2015. Jean Claude Byungura, Henrik Hansson, Thashmee Karunaratne. Expanding Learning Scenarios - Opening Out the Educational Landscape, 64-64
  • 2014. Naghmeh Aghaee (et al.). DSV writers hut 2014, 33-40
  • 2014. Thashmee M. Karunaratne (et al.).

    Learning from graphs has become a popular research area due to the ubiquity of graph data representing web pages, molecules, social networks, protein interaction networks etc. However, standard graph learning approaches are often challenged by the computational cost involved in the learning process, due to the richness of the representation. Attempts made to improve their efficiency are often associated with the risk of degrading the performance of the predictive models, creating tradeoffs between the efficiency and effectiveness of the learning. Such a situation is analogous to an optimization problem with two objectives, efficiency and effectiveness, where improving one objective without the other objective being worse off is a better solution, called a Pareto improvement. In this thesis, it is investigated how to improve the efficiency and effectiveness of learning from graph data using pattern mining methods. Two objectives are set where one concerns how to improve the efficiency of pattern mining without reducing the predictive performance of the learning models, and the other objective concerns how to improve predictive performance without increasing the complexity of pattern mining. The employed research method mainly follows a design science approach, including the development and evaluation of artifacts. The contributions of this thesis include a data representation language that can be characterized as a form in between sequences and itemsets, where the graph information is embedded within items. Several studies, each of which look for Pareto improvements in efficiency and effectiveness are conducted using sets of small graphs. Summarizing the findings, some of the proposed methods, namely maximal frequent itemset mining and constraint based itemset mining, result in a dramatically increased efficiency of learning, without decreasing the predictive performance of the resulting models. It is also shown that additional background knowledge can be used to enhance the performance of the predictive models, without increasing the complexity of the graphs.

  • 2013. Thashmee Karunaratne, Henrik Boström, Ulf Norinder. Intelligent Data Analysis 17 (2), 327-341

    Quantitative structure-activity relationship (QSAR) models have gained popularity in the pharmaceutical industry due to their potential to substantially decrease drug development costs by reducing expensive laboratory and clinical tests. QSAR modeling consists of two fundamental steps, namely, descriptor discovery and model building. Descriptor discovery methods are either based on chemical domain knowledge or purely data-driven. The former, chemoinformatics-based, and the latter, substructures-based, methods for QSAR modeling, have been developed quite independently. As a consequence, evaluations involving both types of descriptor discovery method are rarely seen. In this study, a comparative analysis of chemoinformatics-based and substructure-based approaches is presented. Two chemoinformatics-based approaches; ECFI and SELMA, are compared to five approaches for substructure discovery; CP, graphSig, MFI, MoFa and SUBDUE, using 18 QSAR datasets. The empirical investigation shows that one of the chemo-informatics-based approaches, ECFI, results in significantly more accurate models compared to all other methods, when used on their own. Results from combining descriptor sets are also presented, showing that the addition of ECFI descriptors to any other descriptor set leads to improved predictive performance for that set, while the use of ECFI descriptors in many cases also can be improved by adding descriptors generated by the other methods.

  • 2012. Thashmee Karunaratne, Henrik Boström. 11th International Conference on Machine Learning and Applications (ICMLA), 409-414

    Standard graph learning approaches are often challenged by the computational cost involved when learning from very large sets of graph data. One approach to overcome this problem is to transform the graphs into less complex structures that can be more efficiently handled. One obvious potential drawback of this approach is that it may degrade predictive performance due to loss of information caused by the transformations. An investigation of the tradeoff between efficiency and effectiveness of graph learning methods is presented, in which state-of-the-art graph mining approaches are compared to representing graphs by itemsets, using frequent itemset mining to discover features to use in prediction models. An empirical evaluation on 18 medicinal chemistry datasets is presented, showing that employing frequent itemset mining results in significant speedups, without sacrificing predictive performance for both classification and regression.

  • 2011. Thashmee Karunaratne. ECML/PKDD 2011: Workshop of Collective Learning and Inference on Structured Data

    The recent studies of pattern mining have given more attention to discovering patterns that are interesting, significant, discriminative and so forth, than simply frequent. Does this imply that the frequent patterns are not useful anymore? In this paper we carry out a survey of frequent pattern mining and, using an empirical study, show how far the frequent pattern mining is useful in building predictive models.

  • 2010. Thashmee Karunaratne, Henrik Boström, Ulf Norinder. Ninth International Conference on Machine Learning and Applications (ICMLA), 2010, 828-833

    Graph propositionalization methods can be used to transform structured and relational data into fixed-length feature vectors, enabling standard machine learning algorithms to be used for generating predictive models. It is however not clear how well different propositionalization methods work in conjunction with different standard machine learning algorithms. Three different graph propositionalization methods are investigated in conjunction with three standard learning algorithms: random forests, support vector machines and nearest neighbor classifiers. An experiment on 21 datasets from the domain of medicinal chemistry shows that the choice of propositionalization method may have a significant impact on the resulting accuracy. The empirical investigation further shows that for datasets from this domain, the use of the maximal frequent item set approach for propositionalization results in the most accurate classifiers, significantly outperforming the two other graph propositionalization methods considered in this study, SUBDUE and MOSS, for all three learning methods.

  • 2009. Thashmee Karunaratne, Henrik Boström. The Eighth International Conference on Machine Learning and Applications, 196-201

    Graph propositionalization methods transform structured and relational data into a fixed-length feature vector format that can be used by standard machine learning methods. However, the choice of propositionalization method may have a significant impact on the performance of the resulting classifier. Six different propositionalization methods are evaluated when used in conjunction with random forests. The empirical evaluation shows that the choice of propositionalization method has a significant impact on the resulting accuracy for structured data sets. The results furthermore show that the maximum frequent itemset approach and a combination of this approach and maximal common substructures turn out to be the most successful propositionalization methods for structured data, each significantly outperforming the four other considered methods.

  • 2008. Thashmee Karunaratne, Henrik Boström. Trends in Intelligent Systems and Computer Engineering, 141-153

    Typical machine learning systems often use a set of previous experiences (examples) to learn concepts, patterns, or relations hidden within the data [1]. Current machine learning approaches are challenged by the growing size of the data repositories and the growing complexity of those data [1, 2]. In order to accommodate the requirement of being able to learn from complex data, several methods have been introduced in the field of machine learning [2]. Based on the way the input and resulting hypotheses are represented, two main categories of such methods exist, namely, logic-based and graph-based methods [3]. The demarcation line between logic- and graph-based methods lies in the differences of their data representation methods, hypothesis formation, and testing as well as the form of the output produced.

    The main purpose of our study is to investigate the effect of incorporating background knowledge into graph learning methods. The ability of graph learning methods to obtain accurate theories with a minimum of background knowledge is of course a desirable property, but not being able to effectively utilize additional knowledge that is available and has been proven important is clearly a disadvantage. Therefore we examine how far additional, already available, background knowledge can be effectively used for increasing the performance of a graph learner. Another contribution of our study is that it establishes a neutral ground to compare classifi- cation accuracies of the two closely related approaches, making it possible to study whether graph learning methods actually would outperform ILP methods if the same background knowledge were utilized [9].

    The rest of this chapter is organized as follows. The next section discusses related work concerning the contribution of background knowledge when learning from complex data. Section 10.3 provides a description of the graph learning method that is used in our study. The experimental setup, empirical evaluation, and the results from the study are described in Sect. 10.4. Finally, Sect. 10.5 provides conclusions from the experiments and points out interesting extensions of the work reported in this study.

  • 2007. Thashmee Karunaratne, Henrik Boström. IMECS 2007, 153-157

    Incorporating background knowledge in the learning process is proven beneficial for numerous applications of logic based learning methods. Yet the effect of background knowledge in graph based learning is not systematically explored. This paper describes and demonstrates the first step in this direction and elaborates on how additional relevant background knowledge could be used to improve the predictive performance of a graph learner. A case study in chemoinformatics is undertaken in this regard in which various types of background knowledge are encoded in graphs that are given as input to a graph learner. It is shown that the type of background knowledge encoded indeed has an effect on the predictive performance, and it is concluded that encoding appropriate background knowledge can be more important than the choice of the graph learning algorithm.

  • 2006. Thashmee Karunaratne, Henrik Boström. Proceedings of World Academy of Science, Engineering and Technology 15, 49-51

    Logic based methods for learning from structured data is limited w.r.t. handling large search spaces, preventing large-sized substructures from being considered by the resulting classifiers. A novel approach to learning from structured data is introduced that employs a structure transformation method, called finger printing, for addressing these limitations. The method, which generates features corresponding to arbitrarily complex substructures, is implemented in a system, called DIFFER. The method is demonstrated to perform comparably to an existing state-of-art method on some benchmark data sets without requiring restrictions on the search space. Furthermore, learning from the union of features generated by finger printing and the previous method outperforms learning from each individual set of features on all benchmark data sets, demonstrating the benefit of developing complementary, rather than competing, methods for structure classification.

  • 2006. Thashmee Karunaratne, Henrik Boström. Proceedings of the Ninth Scandinavian Conference on Artificial Intelligence (SCAI 2006), 120-126
  • 2006. Thashmee Karunaratne, Henrik Boström. Proceedings of the Second IASTED International Conference on Computational Intelligence

    Existing methods for learning from structured data are limited with respect to handling large or isolated substructures and also impose constraints on search depth and induced structure length. An approach to learning from structured data using a graph based propositionalization method, called finger printing, is introduced that addresses the limitations of current methods. The method is implemented in a system called DIFFER, which is demonstrated to compare favorable to existing state-of-art methods on some benchmark data sets. It is shown that further improvements can be obtained by combining the features generated by finger printing with features generated by previous methods.

Visa alla publikationer av Thashmee Karunaratne vid Stockholms universitet

Senast uppdaterad: 10 juni 2019

Bokmärk och dela Tipsa