Understanding program behavior is challenging for novice learners. To address this problem, various program visualization systems such as TEDViT have been developed. TEDViT allows teachers to establish drawing rules, ensuring that their intentions are reflected in the visualization results. However, visualizing programs edited by learners using TEDViT has been hindered by its rule notation. In this study, we propose a novel rule designed to accommodate learner-edited programs. Furthermore, we develop a new system based on this rule.

In second language learning, learners learn four skills: reading, writing, listening, and speaking. However, compared with the other three skills, speaking skills have rarely been studied in environments wherein they can be learned. Therefore, we proposed a method for recognizing Japanese word accents using the pitch average of the whole word and that of each mora as part of speaking skill learning. We demonstrated that this method outperforms the pitch drop detection method. Using the proposed accent recognition method, we embedded a function that enabled learners to learn the accents of target words into the feedback phase of an existing Japanese language learning support system. This allowed learners to speak the target words and receive feedback on whether their accents were correct.

Several studies have been conducted to improve program comprehension through code tracing. However, most of these studies have focused on a single flow of control, with insufficient proposals to support program comprehension of multiple flows of control. It is difficult for novices to trace code with multiple flows in the correct order. In this study, we propose a code trace support environment for letting learners to trace the appropriate flow in execution with synchronization constraints. It also supports the component-based comprehension of synchronization mechanisms in multiple flows. The evaluation experiment confirmed the learning effect of understanding the appropriate order of steps. However, several issues regarding the component-based comprehension have been identified.

This paper describes a learning support system that allows learners to answer code fragments fitting blanks in target program code while observing program visualization (PV), and an evaluation experiment conducted to measure the effectiveness of the system. Most existing PV systems have functions to visualize changes in target domain world caused by each execution of statement and to allow learners to answer questions provided by the system, which is not sufficient in terms of learner engagement. In this study, we extended the PV system developed in our previous study by adding a function to provide fill-in-blank exercise to learners and allow learners to answer it, and a function to visualize the PV according to the learner's answer. The effectiveness of our system was measured by conducting a comparison experiment with a previous system. Although the sample size was not sufficient, the experimental results suggest that learning with our system could contribute to improve learners' understanding of program and that our system would be more effective than the previous system.

This study portrays a learning support system that visually represents the structure of stepwise algorithm refinement, also known as the structure of program code abstraction, aligned with the teacher's instructional objectives. We focus on the varying needs of learners in programming education and consider the possibility that different levels of stepwise refinement may be required based on their knowledge and experience. Novice learners with limited programming experience may need program code abstractions in finer steps, rather than the abstraction of numerous statements simultaneously. For learners with a certain level of experience, abstraction with excessive steps may hinder their holistic understanding of the entire program from a higher perspective. In this study, we developed a program visualization system that empowers teachers to customize the visualizations of the structure of stepwise abstractions and the data structures utilized by the target program according to their instructional intentions. We introduced the proposed system into an actual class and evaluated its effectiveness by measuring the degree of improvement in learners' understanding based on tests.

In programming exercise classes, teachers or teaching assistants (hereinafter referred to as “TAs”) circulate among the students to answer their questions. However, it is difficult to detect student s who reach impasses, to instantly identify the part of the program in which students reach impasses, or to grasp the impasse status of the entire class. To solve these problems, we have developed a system with the ability to display impasses detected from students’ exercise information on the source code and to aggregate and display the impasse for the entire class.

TEDViT is a program behavior visualization system that can reflect explanatory intentions of a teacher by visualizing it in a target world. This system can dynamically reflect changes in variables caused by program behavior in terms of numerical values on objects. However, it is difficult to map the variable values to the position, size, and color of the object. In this study, we implemented a function in TEDViT to dynamically map variable values to object properties. We report an experimental evaluation using 10 subjects.

In general, college students enrolled in computer science courses are provided some software design classes and exercises. In a cross-reviewing learning/exercise environment, however, beginners often cannot provide useful feedback and viewpoints to others. One of the reasons for this is that beginners are not confident in their understanding of the software design models. These learners should be encouraged by strengthening their understanding of the models. In this research, we propose a self-review supporting environment that provides learners with feedback generated by translating their Unified Modeling Language (UML) class diagrams into natural language descriptions. This enables beginners to recognize discrepancies between what they would like to describe in their class diagrams and what they actually describe in their class diagrams by comparing them in natural language. We designed three types of templates to make learners aware of the discrepancies. Moreover, we performed an experiment to evaluate whether the feedback produced by translating a learner's model into natural language can support the learner to detect errors in the model themselves, and whether this feedback can support the learner to correct errors in the model themselves. The results of the experiment indicated that the subjects detected errors in their class diagrams with 89.2% accuracy and detected 47.3% of all errors in their diagrams. Furthermore, the subjects corrected the errors they correctly detected with 78.6% accuracy.

SQL injections lead to serious damage such as leakage of personal information and are prevalent online. Web Application Firewalls (WAFs) are used to detect SQL injections, but its detection is not perfect. Therefore, to validate detected attacks, a more detailed manual analysis is required. Furthermore, due to confidentiality and availability, it is often impossible to obtain information on monitored web servers. In this paper, we propose a method to identify the intention of SQL injections in order to shorten the analysis time in the situation where limited information is available. The proposed method analyzes only a partial SQL query in an HTTP request and estimates its intention by syntax analysis and emulating. Evaluation results revealed that the proposed method was able to correctly identify the intention with an accuracy of 83.1% and 71.9% for an artificial dataset and a real-world dataset respectively.

In the field of programming education, learners often cannot fully understand the behavior of algorithms and programs, because it is difficult for them to capture a realistic image of their behavior. Many researchers have proposed methods to visualize “the behavior of algorithm programs” as a solution to address this problem. Many programming learning support systems that incorporate these methods. We have developed a programming learning support system, TEDViT, that allows teachers to control the method of visualization. However, TEDViT is a programming language compatible with the C programming language, hence, the system cannot handle visualization in object-oriented languages. We developed and evaluated a prototype system that can visualize the proposed model.

In this paper, we describe a learning support system for software component design for learners at the early programming learning stage. In general, the source code used in instructional programming exercises does not have a testing code and is not designed to rapidly verify behaviors. We think one of the difficulties for students in designing their source code to be testable is that most cannot analyze their component design themselves. Therefore, we summarized the testability features applicable to such learners and designed our learning support system to improve instructional exercises in analyzing source codes and creating improved component design through feedback. The learning outcomes using the system were evaluated based on the subjects’ improved scores between pre- and post-tests on analyzing source codes and creating improved component design. The evaluation results indicated the effectiveness of the program with some testability features. We implemented the system in three different programming classes to evaluate the applicability of the system.

In this paper, we describe an extension of TEDViT, the program visualization system that can reflect teachers’ intent of instruction to support learning dynamic data structures. We extend TEDViT to enable the definition of drawing rules using values of target program variables and pointers corresponding to the execution status when the rule is fired. The evaluation results reveal that our extension could significantly reduce the amount of drawing rules and the time required for defining program visualizations. It suggests that our approach would support program visualizations with dynamic data structures to a certain degree.

Dictogloss is one of the methods of learning a second language. In dictogloss activity, learners listen to a short text spoken by the teacher, after which they first reconstructs what they have heard individually and then complete the text in discussion with one another. Having already implemented a Japanese dictogloss learning environment without learning partners and a teacher for the beginner level. We focus in this paper on the learning environment for the intermediate learner. We anticipate that the learning by teaching method will help intermediate learners to improve their target knowledge. We have thus implemented teachable agent in our previous environment.

In programming classes, teachers and teaching assistants (TAs) cooperate to support students’ programming exercises. A teacher is required to support many students and sometimes hand them over to another teacher or assistant. However, it is difficult to check students’ source code without interrupting the students, to share tutoring information among teachers and TAs, and to reuse tutoring documents created for one student. In this paper, we propose a tutoring support system based on reusable tutoring content and semi-automatic assessment of students’ code and report on a trial conducted in university programming classes that found the system could spontaneously promote tutoring content sharing and reuse.

TEDViT is a program visualization tool that allows teachers to apply an intention of description (IOD). Using this tool, teachers write a rule set to apply IOD. However, it takes a relatively long time to write the rule set. In this paper, we measure the amount of time required to write rule sets by conventional interface. We determine the reasons why it takes long time and suggest solutions. Thus, we constructed a supporting system that has features corresponding to these solutions. Our experimental results show that the system reduces total writing and debugging time by 41%.

In our previous research, we developed a learning support system for teaching a systematic debugging process. However, it was observed that all the subjects were unable to learn the entire process but some managed to learn it partially at experimental evaluation. We considered following two reasons: (1) the level of debugging skills we expected for subjects was higher than their actual level and (2) subjects were not offered lectures about the debugging process and required skills but were provided with only some exercises. For (1), we consider that they need to gain more basic debugging skills. For example, we decided to teach a skill to observe variables’ values. For (2), we designed a lecture that learners could attend to learn the process and debugging skills. The lecture has instruction and exercise parts. Next, we extended our learning support system to be adapted to the lecture and assist learners who face difficulties in the exercises. In this paper, we report a design of the learning support system and the lecture.

In this paper, we describe a learning support system that can visualize program behavior in memory image and status of the target domain world, and classroom practice that introduces the system to a real class. Several learning support systems have been developed for supporting novice learners in understanding program codes. However, visualization in existing systems often conceals the concrete value of variables such as pointers; the way in which each visualized object is located on the memory is not made explicit. In order to address this issue, we developed a learning support system that visualizes simultaneously and synchronously the memory image that is the field that presents the concrete value of variables, and the target domain world that is the field that presents logically the data structures processed by the program. Moreover, in order to allow teachers to set their instruction content based on the growing variety of learner background knowledge, our system visualizes the status of the target domain world according to the visualization policy defined by the teacher. We conducted classroom practice for understanding pointers in connection with a memory model, thus introducing our system to a real class. In this practice, we included a learning step for the students to observe the memory image of variables and compare it with the status of the target domain world, which had not been included in our preceding practice. Analysis of answered scores in a questionnaire conducted after the practice suggests that our system contributed to reducing irregularities in the participants’ understanding, and that appropriate learning support reached every participant of our practice using our system.

In this paper, we describe how to improve our Japanese dictogloss system to realize richer dialogue to pseudo learner’s partner (called learner agent) and more usable operation using the system on the study. Our existing dictogloss system has some functions supporting learners’ self-study with dictogloss activities. In preliminary evaluation of the existing system, we got the result that indicate the effectiveness using our dictogloss system. Finally, we report a simple evaluation results.

A learning environment where human learners learn by teaching computer agents generally has some advantages for their effective learning. One of the expected advantages is the learners can be aware of the related knowledge and the strategy of problem solving. However some human learners cannot succeed in teaching the agents as a learner even if these learners can completely resolve the problems. It is because these learners cannot diagnose the understanding level of the agents. Even if these learners can teach the answer of the problems, they cannot identify the causes of the agents’ mistakes and lead them to the correct solution. In this paper, to support their learning by teaching, we propose a new learning environment where human learners learn by teaching as a teaching assistant. In this environment, these learners can receive some assistance from the teacher agent. The teacher agent supports the learners in diagnosing the causes of mistakes based on the related knowledge and the strategy of the problem solving, when the learners become stuck. Our preliminary evaluation shows the advantage to enable learners to be aware of the related knowledge and strategy.

In this paper, we describe the code reading environment according to visualizing both variable’s memory image and target world’s status. We had constructed those environment that have two important function. One is the function by which the teacher can reflect his/her intention for instruction on the view of target world’s status. Another is the function by which the programming beginner can more deeply understand a program using our proposed environment.

In this paper, we describe a code reading support environment and a classroom practice using our system for understanding of nested loop. In our preceding work, we had practiced an exercise class in nested loop with a code reading support environment. The evaluation results suggested that students could obtain an expected learning effect roughly by using our system. However, we also found some of them had reached a learning impasse in the classroom. We tried to cope with them, based on two supporting approaches; to bridge a gap between generalization structures of program code and corresponding operations, and to make students capable of predicting the behavior of the nested loop. We extended the preceding system with some new functions according to our approaches, and practiced renewed exercise class in nested loop with our system. The evaluation results suggest that our new system has a certain correlation with understanding of nested loop.

In this paper, we describe an algorithm and programming learning environment by illustrating the relationships among programs, target domains, and operations. Our proposed learning environment supports software programming learners in understanding code that includes nested loops by using a function that visualizes the differences in the target domain's state before and after executing a sequence of operations. We evaluate this environment in the classroom, and the results suggest that the proposed environment improves the understanding of programming beginners.

In this paper, we describe how to improve a Japanese language education environment for nonnative speakers (dictogloss system oriented for focus on form) and its evaluation. Our existing dictogloss system already has some functions supporting learners' self-study with dictog loss activities. However it has not been evaluated in practical situations. To adapt this system to real foreign language education scenes, we improved the system so that it accepts Roman alphabet input without kana-kanji conversion, which some learners actually prefer to input with kana-kanji conversion. We also developed a recording function of a learner's operations which enables us to analyze activities of actual learners. Preliminary evaluation of our improved dictogloss system with actual learners shows that our improved dictogloss system has better learning effect than a conventional dictation environment.

In this paper, we proposed novel learning support environment for understanding programs and algorithm by emphasizing three worlds: programs world, domain world and operations world. To lead learners deeply understand programs and algorithm, we suggest "guidance feature" and "externalization support feature" in learning environment. Hereby, we guess that learners could bridge their gap between programs world and operations world by these features.

In this paper, we discuss how to develop a dictogloss system oriented for focus-on-form (FonF) instruction. Dictogloss is a multiple skills collaborative activity in which learners listen to a text and try to reconstruct the text in collaboration with their partners. Focus on form is a pedagogical approach aiming at improving learners’ ability to produce grammatically correct sentences. A dictogloss system incorporating FonF instruction would therefore be of great help for improving learners’ correct understanding and production of their target languages. We discuss what functions are necessary for a dictogloss system and how to implement the functions. Preliminary evaluation of our system shows that the system can act as a dictogloss partner for second/foreign language learners.

We propose a method to construct an educational system that supports learners to relate their understanding on concrete operations to both generalized algorithm and program codes. The system has functions visualizing relations among program codes, a sequence of operations achieving the purpose of the target algorithm, and status of the domain world. We implement the system and perform preliminary experimental evaluation. The result suggests that our system has some educational effects as we expect.

We have constructed a learning environment in which algorithm/programming learners can operate objects on a linked list model in its domain world. The main advantage of our system is that our model reflects an expert programmers’ perspective: some concepts of programming, and some implicit restrictions caused by those algorithms are written for computers not for people. Such a perspective is important for understanding algorithms; however, it is not common that this perspective is taught to learners explicitly. In this paper, we proposed such a model for a linked list as an example of dynamic data structure. An evaluation of the system and model with the programmers’ perspective showed a marginally positive effect for understanding algorithms.

In a pedagogical approach called focus on form (FonF), a teacher lets learners focus their attention on using grammatically correct forms. However the learners may use forms different from targeted linguistic forms (forms that the teacher intends to teach). This study aims at improving FonF instruction in our system in such a way that the system guides learners to use targeted forms. In guiding the learners, the system behaves differently in accordance relative necessities of forms in a task.

This is a report of the Embedded System Architect Developing (ESAD) program that is currently being conducted in Hamamatsu city, Shizuoka prefecture in Japan. The program has been designed for limited experienced practitioners who expect to be Embedded System Architect (ESA) in the future. In an intensive retreat environment, this program provides future ESAs with a project-based curriculum that covers software engineering, process control systems, and systems engineering. The aim of the program is not only to encourage a deeper understanding of the relevant technology, but also to promote the development of an inter-industry engineers' community in the Hamamatsu area. The program has been conducted three times during 2008 and 2009, with approximately twelve future ESAs participating in each cycle. The beneficial effects of ESAD have been confirmed by observation and feedback from the program participants.

A language education system oriented for focus-on-form instruction should recognize which linguistic forms are focused in a particular lesson, and it should also evaluate how properly the focused forms are used in learners' utterances. This paper discusses how forms should be described in the system, how the form description helps the system recognize focused forms, and how the system evaluates learners' use of the focused forms. Preliminary evaluation of the system shows that the system correctly detects focused forms and that it successfully evaluates the use of the detected focused forms. Neither incorrect detection nor evaluation failure has been attested in its preliminary evaluation.

Presents a design of a collision avoidance switch for the collision avoidance multiple broadcast (CAMB) tree using transistor-transistor logic (TTL) devices. A simulation study exploring the performance of the broadcast star network in both synchronous and asynchronous operations is also presented. Simulation results show that the slotted broadcast star achieves better performance than the unslotted broadcast star, especially when the traffic load is heavy.>