Categorized Questions Dataset QuBAN

This dataset was made for the "A Query-Based Approach to Mitigate the Shortcomings of Widely Used Learning Methods through E-Learning" paper. This Dataset's Version 4 uses an updated algorithm for evaluating questions compared to the previous versions and has all the files separated into different csv files instead of excel sheets.

This dataset consists of 4 .csv files -

Data_Structure.csv

2. Introduction_to_Computers_and_Research.csv

3. Irrelevant_Questions.csv.

4. Blooms_Taxonomy.csv

File-1 consists of the scored and categorized questions from the "Data Structure" course.

File-2 consists of the scored and categorized questions from the "Introduction to Computers and Research" course.

File-3 consists of the irrelevant questions which do not belong to the courses above but were asked by the students from those courses.

File-4 consists of the keywords of Bloom's Taxonomy used to evaluate the questions in this dataset.

The first 3 files consists of questions asked by students of Independent University, Bangladesh on the Summer 2023 Semester in Computer Science Courses. The last file contains the Bloom's Taxonomy keywords that were used for the question evaluation.

This dataset was created to highlight the usage of AI in educational context. It particularly focuses on questions as our goal was to revive the innate curiosity of students to learn through querying. Computer science (CS) courses have been particularly focused as we found higher dropout rates in CS courses compared to others.

The questions in this dataset have been manually pre-processed and categorized according to their course and topics. They have also been scored using Bloom's taxonomy's six levels of questions [remember (10 points), understand (20 points), apply (20points), analyze (25 points), evaluate (35 points), create (40 points)]. Any question above 100 points is considered a high level question and the maximum attainable score for a single question is 150 points.