MPDizon, RAJDeluao,MJGMispeñas,HLTToyorada, RSGamboa
In College, programming subject is an introductory course in any information technology related programs. However, in most cases, novice students had a hard time in this subject and often failed to pass because they find difficulty in doing their laboratory exercises. Some identified reasons to these are the lack of the following: problem solving ability, knowledge of the programming language and understanding of the application domain, conceptualization of the execution of the program and assistance, assistance coming from their instructors and from the programming environment itself. These programming environments returned undecipherable compilation error messages which brought additional burden to the novice programmer in understanding its meaning and affected their way of providing solutions, thus leading to failure in submitting their laboratory exercises.
In this study, the researchers applied data mining in extracting information from students’ DevC++ Compilation Log. Dev C++ IDE is modified to capture students’ compilation logs, written in text files and forwarded to the server. Text files containing log information are sent to Dev C++ Log Miner; these were analyzed to generate visualization and interpretation of results. Visualizations and reports generated reliable information such as error count per compile, committed error count, compile time interval and character count per compile to determine novice students’ programming performance and challenges. Instructors can now thoroughly assess students, and help them improve their programming skills.