This Research to Practice Full Paper presents in this paper. Finding logical errors is the most difficult skill for students from all sorts of disciplines that involve computer programming. Unfortunately, because of this difficulty, some students resort to plagiarism. Plagiarism corrupts the evaluation process. Several tools are used in different researches for this purpose. Recent research has defined a taxonomy of the most relevant types of plagiarism that can be found in source codes. The Hybrid Framework was made to map the student plagiarisms using NLP tecnhics and software test automation techniques for automatic exercise correction with tools available for this purpose. This framework has been tested in the laboratory with promising results.

Teaching (and also learning) software architecture is often challenging since it requires a high degree of abstraction from the students besides talent and creativity of the instructors. Given that context, flipped classroom (FC) has emerged as an approach with potential to reinforce software architecture learning. FC learning promotes a greater immersion in a topic than traditional teaching, improving student engagement, interaction and cooperation by offering students the content before the classroom. Despite the FC popularity and its potential to enhance software architecture learning, studies on the use of it in software architecture teaching are still scarce. The main contribution of this paper is then presenting results on the adoption of FC applied to software architecture teaching. A case study involving a lesson on software architecture inside a course on Internet of Things was taught to a control group (19 people taught using traditional method) and to a treatment group (54 people taught using flipped classroom). Results reveal that the treatment group (who received the intervention via the FC), had a significant better (i) assimilation of content (73.1% of the students acquired most of the content according to the applied assessment), (ii) easiness of abstraction (81.2% of students declared that understood the essential concepts), and also (iii) satisfaction with the technique (78.07% of the students were satisfied about the approach). Conversely, the application of traditional method revealed (i) a lower degree of content assimilation (only 19.1% retained most of the applied content), (ii) only 39.9 % of the students understood the fundamental concepts, and (iii) only 25.87% of the students were satisfied and feeling confident about the content learned.

The aim of this work is twofold. First, we identify web accessibility requirements for deaf people who communicate using Sign Language. Second, we define automation approaches for accessibility testing according to the requirements identified. The requirements were identified through a literature review that considered laws, standards, guides, and scientific articles. This review showed the lack of tools that automate the accessibility test for the Deaf. Thus, we propose two approaches to automating accessibility testing: one that requires the programmer to encode metadata for accessibility testing, and another that relies only on descriptive analysis of data from other web pages accessible to the deaf. Our work was based on the analysis of 150 sites, 100 of them with metadata analysis, and 50 with descriptive analysis. We discuss the tradeoff for each approach.

Computer science is pervasive, being present in daily life. Computer science students need to develop skills to solve computational problems involving real-world applications. The study presents an exploratory case study that examined the perceptions of computer science students towards real-world problem solving. The participants were 72 computer science graduate students in a human-computer interaction course during two semesters. The students were immersed in a real-world problem solving educational situation. After, a quantitative study was conducted, with data obtained from a questionnaire. Data from the questionnaire revealed that the perspectives of computer science students concerning real-world problem solving are positive. The students recognized that the tasks were valuable, did not present difficulties, were engaged, had fun, attributed long-term importance, judged that the tasks fit their personal development, and it was worth to devote their time and resources. In summary, students in our samples hold the opinion that real-world problem solving provides a meaningful learning experience. The findings provide new insights into the opportunities of the application of real-world problem solving in computer science education.

This Full Paper presents a methodology for classification and analysis of test questions by domain content. Today, performance indicators are used to assess the quality of education. Often, these indicators are obtained through large scale tests applied by public assessment organizations, and their results contribute to the quality analysis of institutions, courses and students. To contribute to a more comprehensive educational assessment analysis, which considers the program content of the test items, we propose an automatic domain classification of exam questions and procedures to report results that allow a better comprehension of student performance based on the domain knowledge represented in the exam. The process starts by automatically classifying the test questions into knowledge domains using dictionaries. Using this classification, the mechanism separates the student's results by domain and generates reports that give a better understanding of the test structure and student outcome by knowledge content. This data can be used by data mining algorithms to predict student outcome in the exam based on their results during the course, allowing preventive actions to be taken to address the problems. A prototype has been implemented and is being used to classify and analyze data from the Computer Science exams of the National Examination of Student Performance (ENADE) - an exam applied in Brazil to assess the quality of all higher education institutions. Results have shown that relevant information can be extracted using this methodology that allows to identify test's domain emphasis, areas where students do better or worst, and how the exam reflects course structure.

Corporate university (CO) become a trend in organizations. However, functional characteristics are little explored by CO for boosting knowledge creation. One of them is functional diversity that has been pointed as a factor to increase innovation, contributing for the development of clear strategies and quick responses to changes at workplace. On the other hand, an appropriate distribution of roles helps to promote individual responsibility and group cohesion. It also contributes to the strengthening of positive interdependence of group members. These aspects are fundamental to the development of cooperative work. In this work, we propose a model for group formation for learning approaching dichotomous functional roles and preferred roles in order to improve the group performance into the courses offered by CO at Court of Justice of Goiás (CJG). From the proposed model, we constructed three group formation algorithms. The first algorithm forms groups randomly (AA). The second is a canonical genetic algorithm (CGA). The third is a hybrid genetic algorithm (HGA). After, we performed a comparative analysis of the results reached by the three algorithms. We observed that the HGA achieves superior results than the CGA and AA.

In this work we developed an algorithm to form collaborative groups on Massive Online Open Courses (Moocs) using Particle Swarm Optimization (PSO) method. Group learning principles are used in this work as an attempt to overcome the dichotomy that exists between the collective, which involves the formation of an online learning community on a massive scale, and the individual, with different interests, prior knowledge and expectations. The proposed PSO algorithm accomplishes the task of forming groups based on two criteria, level of knowledge and interest, thus forming groups with different levels and similar interests, providing better students' interactions and knowledge construction. Results of computational tests showed that the algorithm can meet the criteria for grouping in a satisfactory computation time and it is more efficient than algortithms for group formation commonly approached in the literature. Computational tests have also shown that the algorithm is robust taking into account various data sets and variations of iterations.

Computer science is a creative field to work, creativity is demanded from industry and very relevant and evident in software design processes. So, teaching creativity is an important endeavor in computer science education. This work highlights the importance of verbal creativity, providing empirical evidence that when students engage in other students' ideas, new knowledge emerge and is deepen. In this work we performed two case studies. In the first one were considered three computer graphics classes. Patterns of productive discourse that describe knowledge advancement have been found when students engaged in other students' ideas, criticizing, comparing, evaluating, selecting ideas, considering different alternatives, and pointing positive and negative outcomes. The second case study was run in six human-computer interaction classes. The online discourse was analyzed by means of the following categories: relevance, importance, novelty, bringing outside knowledge to bear a problem, ambiguities, linking ideas and interpretation, justification, critical assessment, and practical utility. The results show that when students are better in categories critical assessment, and linking ideas and interpretation, they are also better in category novelty and also the products generated are better. When they engaged in others' ideas they elaborated better interface designs.

The consistent use of metaphors in the context of Human Computer Iteration (HCI) helps to reduce the cognitive load necessary for understanding the functionality of a computational interface. Students' understanding about a good usage of metaphors in HCI improves their ability to properly critique and design computer interfaces. However, is not simple to students to learn how to use metaphors in interaction design. To teach students to develop interactive experiences by means of metaphors is not an easy task. This work proposes a teaching method for using metaphors during the design of website, desktop, mobile or tablet interfaces and present results of a successful case study concerning the proposed teaching method. The teaching method developed is a model of collaborative learning based on King's questioning model and Ferreira's creative dimensions, that consists in creative tasks allied to questionnaires with structured questions and are designed to encourage interaction, group learning and to foment creativity in students.

The courses in computing, especially those focused on software development, require many skills of novice students of computer programming. Due to the fact that computer science is a new area for most of novice students, they demonstrate great difficulty in initial courses in this area. They do not have specific skills in computer programming. This work presents the results obtained from the application of a teaching method, for computer programming students of the integrated high school mode, that combines patterns-based programming and games. The games uses concrete materials and collaborative activities based on sports in non-formal spaces. The results show improvements in students' motivation and in programming abilities. The method proposed in this work takes advantage of analogies between programming logic and games using concrete materials in order to develop programming skills.

Building written feedback, pedagogically sound, standardized and flexible enough to accommodate students who may be in different stages and learning curves is a complex and laborious task. In this paper, we describe a multinational case study involving diverse types of pedagogical feedback provided to Portuguese and Brazilian novice programming students. Programming errors, especially logical ones, can be used as a consistent metric for assessing learning. The research done looks for an innovative form to define content of several types of feedback. It also aims to create an efficient method for the discovery and mapping of students' logical programming errors. The results obtained so far using this approach are presented and analyzed.

Computer programming is a creative activity. However, while computer scientists have devoted their work to solving complex problems and developing new technology, few have written on the creative process behind their innovations. This work explores programming anti-patterns as beneficial detour to the creation of good programming designs. Examples of programming anti-patterns are: failure to combine linear traversal and a flag variable to search a given value in a uni-dimensional array; to draw wrong analogies from natural language; not consider uncommon cases in a solution; to overlook minor parts of the problem; to assume that a program construct will work in the same manner in all situations; not be flexible in understanding new code, not being aware that there are many different right solutions to a problem. In this research, the teacher of introductory programming course promotes discussions among students involving the bad and good features of program anti-patterns, when there are programming anti-patterns code examples and discussions based on situations having the potential to prevent programming anti-patterns, when programming anti-patterns describe a lack of programming skill. An initial experiment showed improvements in creation of good program designs by students, when the teacher boosts discussions exploring program anti-pattens.