A Semi-Automatic Knowledge Discovery Tool to Identify and Visualize Course Bottlenecks
Bibliographic record
Abstract
Contribution: In this study, we examine the notion of course bottlenecks and their prevalence in various academic courses, particularly those in which a higher-than-average number of students experience failure. Subsequently, we delve into the features of a knowledge discovery software tool that has the capability to identify course bottlenecks and present the findings through a user-friendly graphical interface, accompanied by comprehensible explanations. Background: In many courses, there exist some topics which many students find difficult to comprehend, which we name as course bottleneck. While top and motivated students self-learn those topics with extra effort, remaining students often try to only get a superficial understanding of those topics or skip them altogether. In the end, they fail the course. When analyzing the topics that might have caused a student to fail a course, the information that becomes available is often anecdotal rather than data driven. Besides, what is worse is that the information generally comes late in a semester, if at all. Course evaluations tend to come too late to be of use to the students who report them, and end of semester grades often do not reflect which areas are problematic for students. In fact, course evaluation provides an overall experience of the course without pinpointing the bottleneck of that course. Learning Management Systems focuses on individual student performance level but does not provide any means to identify course bottlenecks. There are many initiatives for identifying curricular bottlenecks, but they do not follow a methodological and understandable approach. In this paper, we fill this void. Intended Outcomes: We discuss a web-based application we have built, which identifies concepts within a course as bottlenecks once given class statistics and course content as input. Such an application will assist teachers and administrators to improve curriculum so that both present and future students are benefited. By reducing curricular bottleneck, this software will make enrolling in a course a better experience and will increase the passing rate thereby reduce attrition. Application Design: By assigning each test question a set of additional tags, which connect it to specific concepts, we are able to track empirical data to identify the concepts which are likely to be bottlenecks. We then consolidate this data with students' grades, consistency, and other factors to provide a numeric score denoting the likelihood for a topic to be course bottleneck. We then present the results with a visual aid to highlight the problematic components in a course in the context of other topics that are covered in that course. Findings: The ongoing implementation of this application at a Midwest University of USA has yielded encouraging outcomes in the identification of course bottlenecks. Currently, our focus lies on transitioning the software from the testing phase to the deployment phase. In addition to refining the core logic of the software, we are actively developing a web platform and graphical user interface (GUI), incorporating features such as access authentication and data entry and storage. Prior to a wider deployment, extensive testing will be conducted across multiple courses on campus. Further research and testing are imperative to enhance and fine-tune the system.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.019 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.012 | 0.004 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.003 | 0.003 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.006 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".