Bridging Syntax and Problem Solving: Educators' Perspectives on Pattern-oriented Intelligent Programming Tutor

This paper presents the development and expert evaluation of PoIPT, a Pattern-Oriented Intelligent Programming Tutor designed to support introductory programming education. PoIPT structures learning content around programming patterns - reusable, idiomatic solutions to common small-scale coding problems - to help students move beyond surface-level knowledge of syntax toward developing robust problem-solving skills. The system integrates instructional content, interactive exercises (including code comprehension tasks, Parsons problems, and code writing), and adaptive support based on individual student performance.A qualitative study involving experienced educators from China, Portugal and Brazil highlights the pedagogical value of programming patterns as conceptual stepping stones that bridge syntax and semantics with higher-level reasoning. Participants affirmed that PoIPT supports student engagement, clarifies core programming concepts, and facilitates progress tracking and motivation through personalized recommendations and visual feedback. The system was also recognized for its adaptability to diverse instructional contexts, such as flipped classroom, supplementary exercises after lectures, and independent learning. Deployment challenges included the need for greater variety in exercises, clearer pattern definitions, and expansion to more advanced programming topics.