Lei Zhang
Assistant Professor
Department of Computer Science
College of Liberal Arts & Science
Course(s) Targeted by this Innovation
CSCI 340: Data Structures and Algorithm Analysis
Development and Implementation
2026-2027
Purpose and Impact
CSCI 340 – Data Structures and Algorithm Analysis is one of the highest-enrollment required courses in the Computer Science department, serving sophomore and junior students each semester across multiple sections. It is a required gateway course for all CS majors and a critical checkpoint for professional readiness.
The course faces two compounding challenges. First, AI tools have made it increasingly easy for students to complete programming assignments without developing genuine understanding — a growing concern across CS education that is difficult to detect and harder to address within a conventional assignment system. Second, NIU's student population is diverse in preparation: students transfer from community colleges and enter from varied academic backgrounds, arriving with highly variable mastery of foundational concepts. A single uniform assignment track cannot diagnose or address these individual gaps. Together, these realities make CSCI 340 an urgent and high-impact candidate for redesign — one whose solution can serve as a scalable model for other high-enrollment courses across the department.
Description of Innovation
This project will design and develop a web-based adaptive learning and assessment system, built on large language model inference — either via cloud API or a locally-hosted open-source model — to supplement the existing auto-grader in CSCI 340. The system functions as an AI-powered mastery platform: after each major topic is covered in lecture, students complete an AI-driven Adaptive Learning Session drawn from a question bank spanning multiple formats, including a new question type introduced by this redesign: code reading, where students analyze, trace, or debug code rather than simply produce it. The system continuously diagnoses each student's knowledge gaps and routes them to questions targeting exactly what they have not yet mastered, with students earning full credit through demonstrated mastery rather than mere submission.
Two capabilities distinguish this system from a conventional auto-grader. First, the adaptive engine targets each student's specific knowledge gaps, making every student's learning path genuinely individualized — particularly valuable given NIU's diverse student population. Second, when a student answers incorrectly, the system generates an immediate explanation addressing their specific misconception. A student who misidentifies hash table complexity receives a different response than one who confuses collision resolution strategies — because the AI diagnoses not just that the answer was wrong, but why. These capabilities establish a reusable AI infrastructure the department can extend to other courses over time.