Abstract
Programming online judges (POJs) are an emerging application scenario in e-learning recommendation areas. Specifically, they are e-learning tools usually used in programming practices for the automatic evaluation of source code developed by students when they are solving programming problems. Usually, they contain a large collection of such problems, to be solved by students at their own personalized pace. The more problems in the POJ the harder the selection of the right problem to solve according to previous users performance, causing information overload and a widespread discouragement. This paper presents a recommendation framework to mitigate this issue by suggesting problems to solve in programming online judges, through the use of fuzzy tools which manage the uncertainty related to this scenario. The evaluation of the proposal uses real data obtained from a programming online judge, and shows that the new approach improves previous recommendation strategies which do not consider uncertainty management in the programming online judge scenarios. Specifically, the best results were obtained for short recommendation lists.
Highlights
Programming online judges (POJs) are e-learning platforms which have been widely accepted in the last few years for programming practices in computer science education and for training in competitive programming scenarios [1,2]
This section reviews in short different concepts about programming online judges, fuzzy tools in recommender systems, and the application of recommender systems in POJ scenarios
The current contribution has been focused on the incorporation of fuzzy logic tools for proposing an approach to recommend problems to solve in POJs scenarios
Summary
Programming online judges (POJs) are e-learning platforms which have been widely accepted in the last few years for programming practices in computer science education and for training in competitive programming scenarios [1,2]. It is currently difficult for several users to find out the appropriate problem for trying to solve, according to their current experience and learning needs This is a typical information overloading scenario, in which users can experiment a high discouragement and frustration when the actual difficult level of the problems they are trying to solve does not match with their current knowledge profile and it is difficult to develop a successful solution. Regarding the e-learning systems research area, it is important to remark that the problem we are facing in the current research is different from the task related to users’ guiding through the learning context, which is usually handled with very traditional approaches such as Intelligent Tutoring Systems or ontologies-supported approaches [14,15] Such approaches require as input an important amount of structured information characterizing e-learning content and user activity associated to the corresponding e-learning scenario, for guaranteeing an appropriate user characterization.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
