Conceptual difficulties and visual representation barriers are often major challenges in learning integers, especially related to understanding negative numbers and mathematical operations such as addition and subtraction. This study aims to identify the problems faced by students in learning integers, evaluate the effectiveness of various teaching methods, and propose innovative approaches to improve students' understanding. The methodology used is a systematic review based on Systematic Literature Review (SLR), by analyzing relevant research articles using PRISMA guidelines. This study examines various learning strategies, such as the use of realistic contexts, innovative visual models, manipulatives, and interactive technology. The results showed that a real-world context- based approach and the use of visual models can significantly improve students' understanding compared to conventional methods. Students who engaged in manipulative-based learning and class discussions also showed better mathematical thinking flexibility. These findings highlight the need for further exploration of the use of real contexts, more innovative visual models, and the application of gamification technology in integer learning. In conclusion, integer learning requires a more innovative approach by integrating realistic contexts, new visual models, and interactive technology. This not only increases student engagement but also supports the development of conceptual and procedural abilities more effectively.

The Linear Programming course, specifically the Transportation Problem topic, is often considered difficult by mathematics education students due to the dense mathematical modeling and procedural algorithms that trigger learning anxiety. This research is driven by the need to address passive classroom dynamics and student heterogeneity through adaptive learning, a method whose implementation is challenging for a lecturer to do alone. This study aims to describe the dynamics of implementing Lesson Study for Learning Community (LSLC) based on an adaptive learning approach to the linear programming transportation problem. A qualitative descriptive methodology was used to capture critical moments in the classroom, dissect thinking obstacles, and document lecturer collaboration. This study focuses on in-depth qualitative observation through data reduction, data display, and triangulation. The findings show that the Plan stage successfully designed an adaptive learning path that integrates sharing tasks (Northwest Corner method) and jumping tasks (VAM with dummy). During the Do stage, this approach effectively activated a caring community, namely refutation followed by constructive feedback. Additionally, students cross-checked directly with the model lecturer to visualize the distribution route flowchart to resolve their bottlenecks. The See stage confirmed that this intervention triggered the active participation of passive students, and identified that minor computational errors were caused by inaccuracies in reading demand capacities. In conclusion, the integration of LSLC and adaptive learning provides an optimal evidence-based ecosystem to accommodate diverse student learning rhythms while continuously improving the quality of mathematics learning