Sagar Sen's Homepage

Life in Flow

Tomas Ruiz Lopez, Cancer Registry of Norway, Oslo

Tomas is a postdoc  who works with me on the SOIGNONS project and the developer of FightHPV. A mobile game to nudge women to attend cervical cancer screening in Norway and Iceland. We also work together in the context of Sweetzpot.

Manoel Ribeiro, Universidade Federal de Minas Gerais, Brazil

Manoel has worked with me to develop a tool called Portinari to explore datasets where patients undergo a sequence of events. The tool is based on graph-databases to store data and a query and visualization system based on flow graphs or Sankey diagrams. He has also worked on machine learning techniques to classify sensor data.


Yuanrui Li, Oslo School of Architecture and Design


Li has worked with me on developing FightHPV and 3D printed prototypes for sensors.

Arnab Sarkar, Insitute of Engineering and Management (Research Trainee 2013 -2014)
Sagar Sen, Stefano Di Alesio, Dusica Marijan, Arnab Sarkar:
Evaluating Reconfiguration Impact in Self-Adaptive Systems - An Approach Based on Combinatorial Interaction Testing. EUROMICRO-SEAA 2015: 250-254
References
Sagar Sen, Carlo Ieva, Arnab Sarkar, Atle Sander, Astrid Grime:
Experience Report: Verifying Data Interaction Coverage to Improve Testing of Data-Intensive Systems: The Norwegian Customs and Excise Case Study. ISSRE 2014: 223-234
Sagar Sen, Jose Luis de la Vara, Arnaud Gotlieb, Arnab Sarkar: Modelling data interaction requirements: A position paper. MoDRE 2013: 50-54
Sagar Sen, Stefano Di Alesio, Dusica Marijan, Arnab Sarkar:
Evaluating Reconfiguration Impact in Self-Adaptive Systems - An Approach Based on Combinatorial Interaction Testing. EUROMICRO-SEAA 2015: 250-254

Masters Students

Anne Silje Løvehaugen, University of New Castle (Master of Science, April 2017- August 2017)

Unsupervised Learning from Lifestyle Factors and Screening History for Women Attending Cervical Cancer Screening in Norway

Waqas Moazzam, Univeristy of Oslo (Master of Science, January 2014- June 2015)

Scientific Hangman: A Framework to Gamify Scientific Evidence for the General Public
Governmental and private funding for research in many fields has resulted in a significant body of scientific evidence. Scientific evidence is made available in the form of thousands of peer-reviewed articles in online digital libraries. This evidence is principally used by researchers, students and on occasions for eventual societal impact such as commercial exploitation and popular science communication. How can we engage the general public in understanding scientific evidence while assuring the provenance from a reliable source? Based on the principles of gamification this thesis addresses the question using a web and mobile framework called Scientific Hangman. Scientific Hangman creates a direct link between scientists and the general public. It contains a web administration tool for scientists to create puzzles and a mobile game based on the traditional game of hangman for general public, adolescents for example, to solve the puzzles in a fun manner. The puzzles in our game are created by researchers through the web administration tool based on their research and an active link to the publication is mandatorily made part of the clue. Players play the game in an  attempt to guess a word given a clue that is simple and informative at the same time. We evaluated our first prototype on a focus group at the Cancer Registry of Norway by 
communicating information from invitation letters for cervical cancer screening. Based on the feedback from focus group we improved the game in design, user experience and interface and added gamification elements to make it fun for the player. We evaluated the improved version to test its perceived ease of use, enjoyment and usefulness by employing the Technology Acceptance Model (TAM). Our results from evaluation of gamified mobile app shows 79.9% and 59.9% agreement on usefulness and enjoyment respectively. However, it shows low agreement of only 42.8% on ease of use feature. On the other hand, evaluation of web administration tool on ease of use and usefulness shows agreement of 82.9% and 80% respectively. We discuss results and also provide a future direction for forthcoming research in gamification of scientific evidence.
References

2. Sen. S, Waqas Moazzam Butt, Trude Andersson, Mari Nygård, Scientific Hangman to Gamify the Understanding of Cervical Cancer Screening Reminder Letters, Conference: 30th International Papillomavirus Conference & Clinical Workshop (HPV 2015), At Lisbon, Portugal

Marek Machnik, Univeristy of Oslo (Master of Science, January 2014- June 2015)

Crowdpinion - Obtaining peoples' momentary opinion in dynamic environments
The main concern of this thesis is finding optimal ways of obtaining information from software users with a special focus on long-term studies on momentary opinion, meaning the short-term opinion that people have at the very moment of asking the questions. I chose the event-contingent protocol of the Experience Sampling Method (ESM) as an efficient methodology for this kind of studies and implemented it in a surveying system called Crowdpinion, which consists of a web panel used by researchers to configure studies and a mobile app for participants of the studies. Long-term ESM studies are in many ways similar to crowdsourced generating of research data and just as the workers in crowdsourcing, the participants in studies need to be motivated in some ways. In this thesis I discuss some motivation factors and describe the gamification elements that I included in Crowdpinion in order to increase participants engagement. Studies on momentary opinion can produce valuable data, which can be combined with results from surveys-based retrospective studies or used separately. Having the data from a momentary Experience Sampling study researchers can learn much more than from retrospective studies. For example, in a classic survey they can ask a question like Did you feel tired at work in last weeks? and obtain responses about what the participants remember about their overall tiredness at work. In an ESM study, the researchers can ask the same question Are you tired now? many times in different situations over long time. By doing this, they obtain much more data, which allows them to analyze how the feeling of tiredness changed in different times of a day, in different days and in relation to different events. It is also important that the responses in momentary studies are not biased by memory, because they are reported almost immediately after analyzed events. The scope of studies that can be conducted using Crowdpinion includes evaluation of public services, User Experience, work environment and many other topics.
References
1. https://www.duo.uio.no/handle/10852/44725

Davlet Dzhakishev, Univeristy of Oslo (Master of Science, January 2013- Dec 2014)

NoSQL Databases in the Enterprise. An Experience with Tomra s Receipt Validation System
The information processing demands of many of today's businesses have outgrown the legacy relational database management system (RDBMS) software resulting from both the data explosive growth and the variety of data models. Today, businesses must manage increasingly large volumes of data that must be available across distributed systems and as well as able to evolve and adapt data models according to the changes of business requirements. Enterprises across all industries are challenged by the task of ensuring scalability of massive quantities of data while at the same time keeping database models simple and flexible at the same time. A new and advanced set of software, "NoSQL" as it is so called, has emerged in response to this challenge and offer new methods for storing data. The NoSQL ecosystem has flourished, with numerous software contributions appearing under the NoSQL umbrella. However, as more enterprises have implemented NoSQL solutions, a distinctive set of criteria has emerged that can help today's IT professionals more easily make use of NoSQL solutions built for enterprise-wide development. In this thesis we investigate the usage of NoSQL solutions in the enterprise environment, where RDBMS traditionally run the show. Tomra AS provided us with the industrial case, featuring implementation of the distributed system for validation of receipts, using two different NoSQL databases. In this work, we describe implementation of the receipt validation system and evaluate NoSQL solutions, based on variety of criteria, such as performance, scalability and ease of use. We also describe the advantages NoSQL approach has over SQL (RDBMS) approaches in a distributed environment. The goal is to help decision makers in the enterprise to make better informed judgements when choosing a particular set of database software for their data handling strategies.
Reference
1. https://www.duo.uio.no/handle/10852/42318?show=full

Santiago Hurtado (Master of Science, July 2010-January 2011)

Subject: Tekio: A Self-Adaptive Vision System

Software that adapts its behavior to an operational context and/or feedback from within is self-adaptive. For instance, a computer vision system to detect people may change its behavior due to change in context such as nightfall and feedback such a movement detected. There is a need to reuse software behavior available in legacy libraries within a self-adaptive framework. This is increasingly important because software must operate under varying operating platforms, resource constraints such as mobile devices, envi- ronmental, social and lingual contexts. Middleware platforms with dynamic component loading aim to aid the creation of self-adaptive systems that can change software behavior at runtime. In this work, we present a self-adaptive system, Tekio, written in Java using the OSGi based middleware platform to load/unload components at runtime. Tekio, provides dynamic adaptation capabilities to a vision system helping it reconfigure given contextual changes. Tekio contains generic components to capture context/feedback, plan an adaptation strategy, and reconfigure domain-specific components. The domain-specific components encapsulate legacy behavior implemented possibly in native languages such as C/C++. In our specific case, we use video processing components, using the OpenCV native library in C. We use Tekio to compose these components in a self-adaptive and reconfigurable video processing chain. We perform experiments to validate that the self- adaptive layer based on OSGi has negligible e↵ects on the throughput despite a high rate of adaptations per second.

Publications

1. Tekio: A Self-Adaptive Vision System, Masters Thesis (Universidad de Los Andes)

2. Reusing Legacy Software in a Self-adaptive Middleware Framework, Middleware 2011

Latest News: Santiago and his friend Boris Perez run a startup called Advance Teknologies in Medellin, Colombia


Ajay Kattepur, IRISA, France (We worked together during the first year of his Ph.D. 2009-2010)   

Subject: Variability Testing in Web Service Orchestrations

Ajay worked with me during the first year of his PhD thesis (his supervisors were Albert Beneveniste and Claude Jard). Ajay was interested in web service orchestrations and I was in variability testing. Hence, we teamed up to study service level agreement variations in web services but exposing them to variations that cover all pairwise interactions in atomic services used in a composite service. The publications below go into more detail with regard to our work.

Publications
Latest News: Postdoc in INRIA-Rocquenquort with Valery Issarny

Teaching

Organization and teaching of industrial courses in Software V&V at SImula Research Labs (will do so every year in SRL since 2013) 

Teaching Assistant for COMP 102: Computers and Computing, McGill University ,Canada (Winter 2007)

Teaching Assistant for Imperative Programming in Java, Ecole des Mines Nantes (Fall 2011)

Mentor for Java Student Project, Ecole des Mines, Nantes (Fall 2011)

(Topic: An MP3 Generator to Support Circuit Training for Rowing)