Discovery Model Based on Analogies for Teaching Computer Programming

Teaching the fundamentals of computer programming in a first course (CS1) is complex activity for professor and also challenge them. Nowadays, there are several teaching strategies dealing with CS1 at university, one which use analogies to support abstraction process that student needs carry appropriation fundamental concepts. This article presents results applying discovery model allowed extraction patterns, linguistic analysis, textual analytics, linked data when using concepts by professors university programs train software developers. For reason, based on machine learning text mining was proposed natural language processing techniques semantic vector space modeling, distributional semantics, generation synthetic data. The carried out nine supervised methods, three unsupervised semi-supervised method involving analysis techniques, main findings showed include keywords, part technical terminology, form verbs statement analogy combine them quantitative contexts neutral or positive phrases, where numerical examples, cooking recipes, games were most used categories. Finally, structure construction teach this validated students.