Number of steps and time duration needed while solving the problem can be considered as a representative of student's learning behavior and their effort of understanding. Learning Analytics is a process for monitoring students' thinking process and can become strong evidence to support teachers' objectives in class. Viat-map is an application that implements the Toulmin Argument method and reconstruction approach. It will assist students to think logically by following the teacher's logic, and it shows a positive contribution in the learning process. The reconstruction process used in Viat-map also offers immediate evaluation of student's mistakes. However, a challenge arises in the Viat-map's evaluation of the student's learning process, as the reconstruction process requires students to have the correct answer to complete the learning process. Having this way, new approaches in using student's behavior as learning process score (Viat-map score) are offered. This study involved 105 Students divided into 58 students from the previous research (2023) and 47 students to confirm the Viat-map score (2024). K-Means clustering was chosen for its ability to perform unsupervised clustering, ANOVA was used for comparing more than two groups, and Pearson Correlation was used to examine the relationship between two sets of data. The results suggest that the number of steps taken and the time spent in the learning process can be considered positive indicators of student behavior in the learning process.

In serial concept mapping, a concept map created during one lecture is extended in subsequent lectures, resulting in a comprehensive map over time. The quality of the initial maps directly influences the quality of future maps, making it crucial to enhance the concept maps created in each lecture. Teachers often develop reference concept maps beforehand to evaluate learners' maps and provide feedback based on these reference standards. Incorporating a reference map sharing phase after individual concept mapping is anticipated to improve the quality of learners' maps in each lecture and positively influence subsequent ones. However, the scratch- building method, where learners freely create all components of the map, poses challenges for effective reference map sharing due to discrepancies between the reference and learners' maps. To address this, we explored the re-composition method, where learners construct concept maps by reorganizing pre-provided components, allowing them to concentrate on structural aspects rather than creating the components themselves. This paper compares the effects of reference map sharing between individual concept mapping using the re-composition method and the scratch- building method. In the re-composition method, components are provided by decomposing the reference map, enabling learners to focus on structural differences during the reference map sharing phase. The study was conducted in university lectures, involving three map-building sessions and two reference map sharing sessions across three lectures for both groups. Results showed that the quality of each map (submap) improved incrementally for the re-composition group but not for the scratch-building group. This suggests that reference map sharing in serial concept mapping is more effective when using the re-composition method.

Teaching mathematics could be a challenging task because the students are expected to understand many interrelated concepts and abstract representations (e.g., mathematical symbols). Some learners, especially those with weaker mathematical foundations, may fail to grasp the relationship between important concepts in the subject. To address such an issue, one of the typical learning strategies is providing structured summary writing (SSW) as an information organizing activity to improve learners' understanding and retention of knowledge through explaining relevant concepts using their own words. However, low-achiever learners who do not have sufficient ability to identify and extract important concepts from the learning materials may have difficulty writing a good summary, thus minimizing their improvement due to a high cognitive load. Despite their potential for enhancing mathematics learning, prior studies have reported some challenges in using summary writing and traditional scratch concept mapping. In this study, kit-build concept mapping (KBCM) is proposed as an alternative to help learners improve their understanding of mathematics through reconstructing a teacher-created concept map from provided components. KBCM scaffolds learners to grasp the structure of the knowledge domain without requiring them to extract the structure from scratch by themselves. The results from an experiment involving 105 undergraduate computer science students from two linear algebra classes confirm the potential of the KBCM to be a better alternative for improving learners' overall understanding of a mathematics subject compared to the typical SSW activity. Some characteristics of KBCM and SSW are discussed.

In this paper, we focus on learning through problem-posing in arithmetic word problems and propose an environment for stepwise learning of problem-posing. As a foundation for stepwise learning, we have summarized the relationship between problem-posing tasks as a problem-posing task graph by setting difficulty levels based on the of conditions that must be considered during problem-posing. This is based on a domain that summarizes the structural conditions necessary for arithmetic word problems to be valid. Based on this, we designed and developed a stepwise learning environment called Monsakun Step. To verify its effectiveness, we conducted an experiment with university students. The results suggest that using Monsakun Step may be particularly effective for learners with a low structural understanding of arithmetic word problems.

Summarization is a conventional task to promote reading comprehension. However, because it is impossible to diagnose the summary immediately and accurately, it is not easy to realize formative feedback for ongoing reading comprehension. To solve this issue, we have proposed Viat-map reconstruction as a task for promotion of reading comprehension. Viat-map is a three-component (Data, Warrant and Claim) representation of the Toulmin Argument. Reconstruction means that a learner is required to reconstruct a Viat-map prepared by an expert by using provided components. By comparing the reconstructed map with the original map, it is possible to realize immediate and accurate diagnosis and ongoing formative feedback for the reading comprehension. We have conducted a comparative experiment of promotion effect for EFL reading comprehension between conventional summarization and Viat-map reconstruction. Fifty-eight second year students of Information Technology Department from State Polytechnic of Malang, Indonesia were involved in the experiments. All of them are using English as Foreign Language. The experiment was conducted in an English course Subject for three weeks. Experiment and control group were created equally based on the initial test of reading comprehension test. The experiment group was using Viat-map, while control group is using conventional summarization method. The result of the experiment was analyzed by using ANOVA statistical analysis and Holm’s Sequentially Rejective Bonferroni Procedure for POST Hoc analysis. The result shows that Viat-map reconstruction overachieved summarization and both groups can maintain the memory one week after the lesson. However, the delay-post test score for experiment group is higher than control group.

Conceptual modelling is a critical ability for a computer science student, enabling them to create a variety of conceptual models such as Entity Relationship Diagrams (ERDs). However, learning to create an ERD is a challenging task especially for novices because it involves various cognitive processes and actions. This research proposes a recomposing approach using kit-build exercise for ERD design learning. In a kit-build exercise, a student is provided a set of components of a structure and asked to recompose the structure using these components. The kit-build exercise has advantages as (1) reducing student’s cognitive load by providing available components and (2) realization of automatic assessment by comparing recomposed structure with the original structure. This exercise has been already implemented for concept maps and the learning effect has also been confirmed both experimentally and practically. Considering the structural similarities between ERDs and concept maps, we have applied the kit-build exercise approach to ERDs. In this paper, we describe the kit- build exercise for ERD with automatic assessment and feedback.

Asynchronous online discussion activity has become a popular form of online collaborative learning (OCL) and has been widely adopted in higher education. In the discussion, learners are expected to effectively co-construct knowledge together through engaging in problem solving discourse. Such an ability to construct and confirm understanding while working together to solve a problem is defined as cognitive presence, which plays a central role in the success of OCL. However, despite various facilitation strategies suggested in prior studies, it is challenging to ensure that learners show sufficient cognitive presence. Providing a preparatory activity that increases engagement with learning materials before participating in the discussion is another strategy that could foster cognitive presence. This study aims to investigate the effectiveness of providing two kinds of knowledge reconstruction activities, namely kit- build concept mapping (KBCM) and summary writing, as a discussion preparatory activity. The results show that using KBCM for reconstructing basic knowledge related to discussion topic has a potential to be better for fostering cognitive presence compared to summary writing. The KBCM activity provides opportunities to engage more with basic knowledge related to the discussion topic, to confirm understanding, and to familiarize themselves with common terms or keywords used in the discussion through reconstructing a teacher-created map. Some directions for future studies are discussed.

Concept map as an assessment tool is widely accepted in the field of education. In this research, assuming the process evaluation of concept map building leads to better evaluation than the product (that is, built concept map) evaluation, we proposed a process evaluation method and conducted an experimental evaluation of the method. The process evaluation method is intended for re-composition type of concept maps, not for general ones. In the re-composition concept map, a learner is provided a set of components of a concept map, that is, links and nodes, and required to recompose a concept map by using the components. The provided components are generated by decomposing the concept map built by an expert as a representation of understanding of the learning target. In this research, we set several submaps with a set of semantics in one expert map and define a good process by recomposing the submap as a series of activities. The process evaluation results were compared with the product evaluation results for the data of a practical use of re-composition of concept map in undergraduate class. As the results of the comparison, we have already confirmed that the result of process evaluation was a better predictor of learner’s understanding than the results of the product evaluation. In this paper as an additional analysis, we reported the analysis of relation between process evaluation and product evaluation. The difference between the evaluation methods could result in a case where there is zero score for process evaluation and full score for product evaluation. However, analysis results confirmed that there is a strong correlation between the two evaluation methods with r value = 0.8. These results suggest that support for the process of submap re-composition is a promising way for learners to recompose a good concept map. We further explore the usability of the process evaluation to the teachers, and it shows a high rate of agreement of the usefulness and appropriateness of the method. Based on the results, we are developing a formative assessment and support environment for concept map re-composition.

The present paper describes the Reciprocal Kit Build approach as a designed activity for collaborative concept mapping. We aimed to investigate the effect of differences in individual knowledge (both prior knowledge and knowledge on task) on knowledge transfer, collaborative product, and learners’ affective responses during co-construction of a concept map with the Reciprocal Kit Build (RKB) approach. We categorized learners into two groups based on their prior knowledge equivalence and their degree of shared knowledge reflected in the individual maps. The RKB approach allowed learners to create an individual map, to reconstruct a concept map from their partner’s map components (nodes and links), and to discuss similarities (or differences) between the initial map with the reconstructed map. The results showed that, following our proposed activity, transfers of individual knowledge regarding the shared and unshared knowledge were considerably high. Although, learners’ differences on prior and shared knowledge did not significantly affect the knowledge transfers and the final collaborative products, different group composition influenced the experiences of learners.

Recently, access to information is easier than before, but most of them are written in English. It leads to another new problem for a non-English speaking country. This research proposes an approach to improve English reading comprehension with the help of Kit-build concept map (KB-map) with an additional function named source connection. A university teacher in Indonesia practically used this approach and the other two approaches in several English classes of the second year of undergraduate student to compare the effectiveness of those approaches. The result shows that KB-map with the source-connection function was more effective in English reading comprehension compare to the traditional summarization and the normal Kit-build method.

Gamification, the use of game elements in non-game contexts, is a popular method to design motivating, enjoyable learning experiences. It is often applied through external rewards like leaderboards and achievements. Research shows that these rewards can negatively affect intrinsic motivation. One way to avoid this issue is by applying the game elements to the internal design of the learning activity. This approach is called Core Gamification. However, there are no concrete methods to apply Core Gamification to learning activities. This paper presents a method for the Core Gamification of learning activities based on information structure manipulation. In this case, information structure manipulation is not done by the code. And they are also not just about the design of the activity. The structures are visible to the user. The manipulation is also done by the user. Case studies with verified learning effects is also shortly presented.

Typically, classroom practices in the collaborative learning context center around three distinguishable levels of activity: individual activities, group work, and whole-class discussion. It is important to identify and analyze the correlation between these levels to enable teachers to understand and improve the dynamics of students’ understanding in the context of collaborative learning. This study proposes a method based on the Kit-Build concept map (KBmap) to analyze the relation between individual activities and group work in the classroom. The KBmap is a type of close-ended concept map that provides decomposed concepts and links from the concept map made by a teacher. This mechanism enables teachers to assess students’ understanding, and facilitate the coordination of learning in the classroom. To evaluate the proposed method, a junior high school in Japan is used as case study.

Conjunctive words connect two sentences logically. This property is divided into same-meaning and different-meaning commutativity. The former emphasizes a meaning that is explicitly represented in the sentences that already exist, but the latter adds a meaning beyond what is explicitly represented in them. Doing exercises to learn how to use both the former and the latter is important because conjunctive words play an important role in reading and writing effective sentences. Exercises to learn the former are easy, as there is only one correct answer to a fill-in-the-blank problem. On the other hand, to demonstrate mastery of the latter, the exercises must be difficult because various correct answers with different meanings exist. In this study, we designed and developed a conjunctive word learning support system for performing conjunctive word exercises to master different-meaning commutativity. In this system, the learner is provided with an exercise that entails assembling a sentence structure composed of two sentences and a conjunctive word and a conjunctive logical structure composed of two explicit situations and an implic it situation. The system also automatically diagnoses the learner’s performance on the exercises and gives feedback based on the diagnosis results. Furthermore, we carried out an experiment to verify the validity of this system, and the results suggest that the system is effective for helping students to master this aspect of language.

It is important to reuse knowledge acquired through problem-solving in programming. To reuse knowledge, it is effective to first understand differences between knowledge items and then to organize that knowledge. Therefore, we propose a method and develop a support system for facilitating knowledge organization in programming. We further examine a model of parts and problem-solving process of parts based on function–behavior– structure aspects. In this paper, we expand a previously developed system to propose an improved system that provides support based on feedback from these models. The aim of this system is to allow learners to consider behavior rather than thinking directly from function to structure.

Kit Build (KB) is a closed-ended and reconstruction concept map that proven positive effects in assessment and learning regarding learners' knowledge understanding. The KB design emphasizes the centralization of teacher understanding that would be reconstructed by learners and ignores their knowledge structure itself. This study focused on a new extension of the KB concept map reconstruction, which is combined with the open-ended style called KB-Mixed. KB-Mixed does not only aim to increase knowledge understanding, but also enhance meaningful learning and facilitate knowledge structure. This study involved university students as participants who were divided into two groups, experimental and control. Students in the experimental group were requested to construct concept maps using the KB-Mixed method, while those in the control group used the open-ended concept map which called Extended Scratch-Build (ESB). The experiment results indicated that KB-Mixed outperformed the control group in terms of knowledge understanding and knowledge structure.

This study investigated a new extension of reconstruction Kit-Build (KB) concept map, which combined with the open-ended style that calls KB-Mixed. To confirm the learners' comprehension of new knowledge, we compared KB-Mixed and open-ended concept map. This study involved 55 2nd-grade university students who were divided into two groups, experimental and control. Students in the experimental group requested to construct concept maps using the KB-Mixed method, while those in the control group using the open-ended concept map, which calls Extended Scratch-Build. The results indicated that KB-Mixed had a significant impact on learners' comprehend, which was represented by the post-test score. Moreover, the concept maps score assessed by expert judgment showed a significant correlation difference to the post-test score.

In a collaborative situation, learners must actively interact with their peers to have meaningful discourse, not only co-present. Researchers have shown that employing a concept map as a representation tool was useful to construct and maintain group shared knowledge, rather than a dialogue only communication setting. The use of concept map positively affected students’ learning outcomes as well as their attitudes. Moreover, prior studies have extended the collaborative concept map activity with individual externalization, concept map sharing and reviewing activity to trigger cognitive change through conflicts. The current study introduces a new extension of collaborative concept mapping activity with Reciprocal Kit Build (RKB). The RKB allows learners in pair to create an individual and collaborative concept map, exchange ideas through reconstruction, and discuss it facilitated by shared and difference maps. Unlike existing studies, our design activity is aimed to promote learning mechanism at boundaries, by boundary-crossing and utilizing boundary objects. This paper explains how the RKB may potentially provoke conceptual changes during collaboration.

It is sometimes challenging for learners of elementary physics to correct their mistaken notions regarding physics concepts. The need to address this phenomenon is illustrated by the existence of common misconceptions such as ‘motion implies a force’ (MIF), in which learners mistakenly believe that force always acts in an object’s direction of motion, as well as educational approaches using error-based simulation (EBS). In a conventional EBS, a learner’s mistake is usually output as a physical phenomenon, but there are cases in which the learner’s mistake cannot actually be shown. For this reason, we propose an EBS in which measurement information is displayed to the learner in the form of a measurement tool together with the physical phenomenon and shows error in response to the learner's mistake, something the conventional EBS is unable to do. In this paper, we propose an EBS learning exercise using a measurement tool in which the learners themselves must determine the measurement in order to confirm the correct answers. Using this method, it is possible for the learner to use trial and error after getting the wrong answer and eventually find the correct answer. It is also possible to support learning in a virtual experiment environment using a computer. Experiments were performed using a system equipped with the measurement tool, and it was verified that the learners made significant progress learning physics through trial and error.

This paper reports a practical use of triangle block model for learning of arithmetic word problems. The triangle block model has been proposed as a bridging model between conceptual representation of a word problem and quantitative representation of its solution. Based on this model, we have developed an interactive environment where a pupil is able to manipulate an integrated representation of the conceptual and quantitative ones as learning of arithmetic word problems. We also designed a series of lessons to use the environment. The lessons were practically conducted for 75 4th grade pupils (in two classes) in an elementary school for 7 class times. As the results, it was confirmed that (1) the pupils and their responsible teacher accepted the lessons as useful ones, and (2) the lessons had learning effect as improvement of structural understanding for the problems.

Kit-building task of a concept map is a promising exercise for strengthening and assessing learner's comprehension for a topic that a learner already has learnt. In order to investigate the value of the kit-building task of a concept map, we are comparing it with multiple-choice task of fill-in-the-blank questions. As a step to realize the comparison, in this paper, it is described an implementation to generate learning task from a series of propositions, that is, (1) kit-building task of concept map and (2) multiple choice task of fill-in-the-blank questions.

Formative assessment is widely used in education area for improve learning achievements. And we propose a framework of Kit-Build concept map for utilizing in ongoing formative assessment. Concept map strategy is already proven to represent learner’s understanding and becomes to a suitable strategy for formative assessment. We investigate the proposition of concept map that is constructed by connecting two concepts via linking words. The framework can gather and assess concept maps as learners’ evidence for identifying learners’ state in classroom situation that is a task of formative assessment. It provides assessment result and diagnosis result to instructor. These results are generated by using the proposition level exact matching method of an automatic concept map assessment. The assessment result is the similarity score between instructor’s concept map (Goal map) and learner’s concept map (Learner map). And the diagnosis result is represented in form of concept map with three error types. The errors show the different proposition between goal map and learners map, which is called individual-goal difference map. Both assessment result and diagnosis result provide information about understanding of each learner on lecture content to instructor that is called individual diagnosis. It can contribute instructor’s feedback that is a key of formative assessment. As more as possible to gather and assess learner’s evidence provides more opportunities to instructor for improving learner’s achievements. Providing overview of learners can reduce analysis time of instructor to recognize learners’ understanding. It means instructor can design and provide instructor’s feedback in less time based on group diagnosis of Kit-Build concept map. The group diagnosis includes group map and group-goal difference map. The group map is generated from overlaying all of learners map. It shows common understanding of learners that is used to generate group-goal difference map, this group-goal difference map shows most common misunderstanding. In this paper, we would like to propose the scenario that is suitable for supporting through of learning process since setting goal, gathering evidence of learners and providing effective information to designing instructor’s feedback for improving teaching and learning in classroom situation.

Learning by problem-posing is promising activity in learning mathematics and considered to contribute the understanding of the problem’s structure. If we can clarify how learners are thinking in steps of problem-posing, we will support learning by problem-posing more. This study investigate what learners think during problem-posing as sentence-integration in terms of intermediate products as well as the posed problems as the resultant product. Problem-posing as sentence-integration defines that arithmetic word problems have a structure and problem-posing is a task to satisfy all the constraints and the requirements to build a valid structure. A previous study shows, in problem-posing of arithmetic word problems as sentence-integration, learners try to satisfy relatively large number of the constraints in the posed problems. This study focuses on the violation of constraints in the intermediate products during problem-posing as sentence-integration. If learners have concern with the structure, they will try to avoid relatively larger number of the violated constraints on the way to the resultant posed problems. We conducted Pearson’s correlation test between the occurrence frequency of intermediate products and the number of violated constraints of them. The result shows the negative correlation between them. We consider that learners had an inclination to avoid as many violated constraints as possible throughout problem-posing activity.

We have developed a learning environment for problem-posing as sentence integration. In this exercise, a learner is required to select and arrange given simple sentence cards for posing problems. On the other hands, learners write problems from scratch in usual problem-posing. Because it is very difficult for learners with reading disability to write a sentence, they cannot exercise by usual problem-posing virtually even though a teacher would like to teach the arithmetic word problems to them through problem-posing exercise. However, several learners with reading disability are able to read a simple sentence. So, we assumed that these learners realize to exercise problem-posing by problem-posing as sentence integration. Moreover, its practical use targeted as sixteen junior high school students in special classroom is reported.

Previous researches on supporting EFL (English as a Foreign Language) reading showed that Kit Build concept mapping (KB-mapping) method has better efficiency than Scratch Build concept mapping (SB-mapping) method in recalling the comprehended information two weeks later. In this research, we set out to investigate the following research question “Why KB-mapping method has better efficiency than SB-mapping method in recalling the comprehended information two weeks later?”. One of the common comprehension styles in the EFL reading comprehension tasks is the sentence-by-sentence, where learners understand the text by closely following each sentence in the text. Learners will be able to sentence-by-sentence answer the Comprehension Test (CT) just after the reading, but many investigations indicated that, this understanding could not be kept for a longer time as is the case in the Delayed CT (DCT). We have a hypothesis that “KB-mapping method promotes a learner to read not sentence-by-sentence, although SB-mapping method does not promote same as KB-mapping method”. In order to proof this hypothesis, we conducted an experiment to compare the learner’s comprehension with KB-mapping and SB-mapping methods. During this experiment, we have monitored the performance of learners in terms of the size and the progress of building learner’s map. In doing so, we added new functions to our system to record the building process of every learner. The outcome of the experiment confirms our previous results regarding the efficiency of KB-mapping. Moreover, the analysis of the progress of map building gives a proof that KB-mapping method helps the learners to avoid the sentence-by-sentence comprehension style. But the SB-mapping method could not help the learners to escape from this style of comprehension. Based on these results, we will design adaptive support for reading comprehension in our future work.

The implementation of learning by problem posing in practical use is facing the issue of inefficiency in time needed for assessment and giving feedback to students’ posed problems. As a solution of this problem, we have developed tablet PC-based software for learning by posing arithmetical word problems named Monsakun. The software is based on Triplet Structure Model of arithmetical word problem. In this research, we investigated elementary school students’ performance in Monsakun. The result shows that students did not pose problems randomly and with some sort of thinking. With further analysis of constraints, we also found that the frequent errors contain few non-meaningful errors and students tried to pose problems satisfying as many constraints required in each assignment as possible. The process of understanding assignment requirements and relating them to suitable sentence cards is an important point especially for young learners to reach deep understanding of the structure of arithmetical word problems.

In solving arithmetic/mathematics word problems, comprehension phase is the most important and difficult phase. In the comprehension phase, conceptual and quantitative comprehension is formed as connection between problem representation and solution representation. Externalization of the comprehension phase is a promising way to support learners to overcome their difficulty in this phase. In this paper, triangle block model is proposed as a way to externalize the conceptual and quantitative comprehension. We have already developed a learning environment based on the model. In this environment, a learner builds the externalized representation of the conceptual and quantitative comprehension by combining provided components. Through an experimental use by seventy-two sixth grade students in an elementary school, we have confirmed that the learning activity designed based on the model is accepted by the students and the responsible teacher of them.

Problem posing is well known as an effective activity to learn problem-solving methods. Although the activity is considered to contribute to understanding of the structure of problems, it is not clear how learners could understand it through the activity. This paper proposes detection of important learner's actions by tracing problem-posing activity sequences. The sequences are considered to represent thinking process of learners and reflect their understanding and misunderstanding about the structure of problems. This paper expects visualizations of the sequences to be helpful to infer learners’ bottlenecks in thinking and misunderstanding behind them. The goal of this paper is to make it possible to analyze problem-posing activity sequence for enhancing the effectiveness in learning. As an example, this paper proposes detection of “trap-state” that is a combination of simple sentences many learners tend to make in problem-posing assignments.

In generally the reading comprehension is a very difficult task for students in all the stages of study, especially when they are reading text in a foreign language, the EFL reading” (English as foreign language) is one of the most common research topics in the learning field. This study aimed to investigate the effect of teaching EFL reading comprehension to Japanese students through Kit-Build concept mapping (KB-mapping) method. In doing so, we have conducted an experiment to compare the student comprehension with KB-mapping and Scratch-Build Concept Mapping (SB-mapping) Methods. In doing so, a comprehension test and delayed comprehension test (two weeks after using) were used. By comparing the results of the Comprehension Test (CT) scores and the Delayed Comprehension Test scores (DCT) for the both conditions groups, we found that the using of KB-mapping had the same effects of SB-mapping in the understanding the text just after the using, but KB-mapping had better effects in the recalling and remembering the text after a while.

We have developed a computer-based learning environment called MONSAKUN to realize learning by problem-posing, where students pose arithmetical word problems by selecting and arranging several sentence cards. As the next step of MONSAKUN development, we analyze the sentence selection process which is considered to reflect students' thinking process. In the first step of analysis, we focused on first sentence selection. We found that the selection approach changed based on different type of story and students’ exercise experience. This result is an important step towards building elaborate process model of problem-posing and adaptive support of the process.

As a support for learners who failed to solve a problem, teaching the way to solve the problem is a general method. However, in order to realize active learning, it is a desirable way to let learners overcome the impasse of the problem solving by themselves. Because it is usually difficult for the learners to do it by themselves, we have proposed “problem simplification” that is a way to find a simplified problem that is included in the original problem from the viewpoint of problem solving. If the simplified problem can be solved by the learner, the difference between the simplified one and the original one is the origin of the impasse. To make clear the origin of the impasse is feasible scaffolding for self-overcoming of impasse. In this paper, firstly we propose a framework to define simplification of problems in elementary mechanics in high school. Then, it is introduced a support system for self-overcoming of impasse of problem solving using the problem simplification targeting elementary mechanics. We also report a practical use of this system in classes of a technical high school.

A framework for indexing problems is proposed, based on which explanation generation and problem sequencing for conceptual understanding in science can be automatized. It helps student learn how to make appropriate models to solve problems and prevents them from superficially read the solution of a worked example and apply it wrongly to others. The results of a preliminary experiment is also described in which the explanations generated based on the framework promoted subjects’ conceptual understanding in mechanics.

Learning by posing arithmetic word problems is well-recognized as an effective means to mastering the use of solution methods; however, several learners cannot pose problems using words. In problem-solving exercises, if a learner cannot understand what is expressed in a problem statement, an illustration of the scene expressed in the statement can help the learner understand. Therefore, in this paper, we propose learning by posing problems with illustrations instead of words. One of the purposes of learning mathematics is to acquire the ability to derive a numerical relation from a scene and to associate a calculation expression with various scenes. We expect that posing problems via illustration and without words is an effective way for beginners to learn. In the problem-posing exercises, learners construct various types of problems. Teachers have to evaluate each posed problem and advise learners on the basis of the evaluations. This task places a heavy burden on teachers. Therefore, we developed a computer-based environment for posing problems using illustrations. The environment provides a learner with a calculation expression and illustrations. The illustrations can be disassembled into several figures. A learner assembles an illustration that describes a story of arithmetic problem that correspond to the calculation expression by combining the figures. The learner uses no sentences in this problem-posing exercise. The environment diagnoses the illustration problem and produces advice to the learner. We conducted a preliminary evaluation of the environment to confirm that a learner who cannot pose problems via words was able to pose problems using illustrations. Our results were promising, indicating that our environment provides a means for learners who cannot pose problems via words.

Problem-posing is effective learning for comprehending problem structure. We have designed and developed the learning environment for problem-posing and performed its practical use for first and second grade students on elementary school continually. The scopes of these systems are one-step addition, subtraction or multiplication arithmetic word problem. The results of these practical uses suggested that the learning environment was effective to comprehend these problem structures. Therefore, as the next step, we have designed and developed a learning environment for posing one-step multiplication or division word problem in order to let learners acquire a difference between multiplication and division. Developed learning environment and its practical use in an elementary school are reported.

This paper proposes “fraction block” as an educational reifications of fraction. Fraction block reifies characteristics of “ratio fraction” as a pair of numerator block and denominator block. The length of the blocks can be changed by keeping the ratio of numerator block to denominator block. By using the fraction block, a quantity represented as a length of tape is derived from another quantity represented as another length of tape. This deriving operation is a reification of multiplication or division with a fraction. We have implemented a learning environment where a learner is able to directly operate the fraction block in order to derive a quantity from a quantity. Experimental use of this environment in an elementary school is also reported in this paper.

We proposed a design method to automatically generate an educational game by substituting the cards of an existing card game for new cards that have practice problems written on them. However, several games in which most players rarely solved problems were created by the method. Therefore, this study examines the reasons why they rarely solved problems. We conducted an experiment that asked test subjects to play useful and useless educational games and their reactions were videotaped. The video recordings and interviews suggest that few subjects solved problem in playing the educational games had no opportunities for solving problems to predict the future game state that would result from each choice to make a favorable choice. On the other hand, all subjects solved problems in playing the educational game had opportunities for the prediction. On the basis of the results, this study also organizes the rules that provides the prediction, and develops a system that detects useless educational games. The results of experimental evaluations of the system suggest that the detection system can detect useless educational game by the detection of the rules for the prediction.

In this paper, an interactive environment for learning by problem-posing targeting arithmetic word problems that are solved by one-step multiplication is described. Its practical use in an elementary school is also reported. We had already developed an interactive environment for learning by posing arithmetic word problems that can be solved by one-step addition or subtraction. Then, a practical use by the first grade students of an elementary school had been performed. The results suggested that the learning by using the environment was effective to improve the student's problem solving performance. As the next step of this previous research, we have developed another learning environment where a learner poses the multiplication word problems. In order to design the environment, we categorize word problems that can be solved by one-step multiplication based on the sorts of quantities consisting of the multiplication. Then, based on the categorization, we have built a task model of the problem-posing. The learning environment is used by a class of the second grade students for 9 class times, and we have confirmed that the scores of problem-posing and problem-solving were improved in the group of students who obtained low scores in the pre-test.

We have already proposed Kit-Build Concept Map (we call it as “KB map”) that is a framework to realize automatic diagnosis of concept maps built by learners and to give feedback to their errors in the maps. In this paper, we report an experimental evaluation of KB map in science learning in an elementary school. Two classes of fourth grade students in an elementary school attended the experiment. “Waxing and Waning of the Moon” in science class was the topic of this experiment. The results suggested that the diagnosis and feedback with the KB map has positive effect to improve learner’s understanding of this topic.

In this paper, we present a design of penalties against the action in contravention of the rule to discourage the tendency of unthinking and useless plays, observed while playing the automatically-generated educational games by our design method. Monitoring player’s activity and penalizing illegal moves are potential ways of preventing unconsidered activities. Because of the illegal moves and adequate penalties are strongly linked in each educational game, it is necessary to design the penalties individually for each educational game, therefore, we investigated the automatic design of penalties for educational games, generated by the EPIC method. We also introduce a function to automatically implement the penalties, and report the experimental evaluations of educational games with penalties.

We have developed an interactive environment for learning by problem-posing targeting arithmetic word problems that can be solved either by addition or subtraction. In our previous work, the environment was used by third or fourth grade students who have already acquired ability to solve the targeted problems. Therefore, problem-posing was an additional practice for the students and the purpose of the learning with the environment was sophistication of their ability. In this paper, practical use of the environment for the first grade students is reported. Just after the classes of problem solving of the arithmetic word problems, the way of problem-posing itself was taught in classes, and then, the environment was used as exercise of the problem-posing. Through this practice, we have confirmed that (1) the first grade students were able to pose problems in the environment, and (2) the practice to pose problems improved their ability not only in problem-posing but also in problem-solving.

In the framework of kit -build conc ept map, (1) a goal map is prepared by a teacher, (2) parts of the map is generated by decomposing the goal map, and then, (3) a learner makes a map (a learner map) by combining the parts. The learner map is diagnosed by comparing with the goal map. The di fferences between the learner map and the goal map should be resolved by the learner to complete the learner ’s map building. In this research , we have investigated a way to support the learner to resolve the dif ferences. Based on the analysis of learner ’s map building behavior, we have proposed three kinds of hints for learners, that is, (I) concept group -centered building, (II) specific concept -centered building, and (III) recently connected concept -centered building . We have implemented a function to gene rate these hints based on the learner ’s building history and conducted a preliminary evaluation of the function .

Abstract.We study about a method to describe knowledge for generating an adequate number of multiple-choice questions. We focus on the fact that instructors can easily explain concrete multiple-choice questions, even if they cannot describe the domain knowledge completely. We propose the Environment of Abstract Rule (EAR), which enables an instructor to describe abstract knowledge by explaining the difference among some concrete multiple-choice questions, and by representing the knowledge in the form of a diagram. In this paper, we describe the concept of EAR and report the results of a preliminary experiment in which we explored whether instructors could describe knowledge easily. As a result, we suggest that the EAR has the potential for generating appropriate knowledge.

Diagnosis of errors in problem solving process is very important to support learner’s problem solving exercise. Although externalization of the problem solving process by a learner is a promising way to realize the diagnosis, this task is hard for a learner and it is very difficult to interpret the expression automatically for a computer system. In this research, we pay special attention to Kit-Build method to help this task. The expression made by the learner is diagnosed by comparing the ideal one, and the differences are targets of learning support. We have design and implemented a learning environment where a learner expresses his/her problem solving process with the provided parts and the expression is diagnosed. In this paper, results of a preliminary use of this environment are also reported.

This paper reports on the practical use of Error-based Simulation (EBS) for learning dynamics in a junior high school. EBS is simulation generated from learners’ erroneous answers. Differences between simulations showing correct phenomena and EBS showing unusual phenomena promote learner awareness of errors. Previously, we implemented EBS for learning about normal reaction forces in statics problems, and confirmed that EBS was effective for learners’ self-correction of errors. In that problem, however, visualization of errors is easy because a correct simulation is always motionless. In this research, we implemented EBS for a dynamics problem, and put it into practical use in a junior high school. Because errors can be visualized as unexpected motion, this is an important generalization of EBS. The learning effect of EBS in dynamics problems was evaluated by comparing the results of pre-tests, post-tests, and delayed post-tests between two classes, a experimental class where students learned with the EBS and a control class where students learned as usual. The results suggest that EBS is also useful in dynamics problems.

We previously implemented an authoring system to generate a learning environment based on a card game expressed in a flowchart . We improved the authoring system to automaticall y generate a learning environment by using words or sentences drawn from the rules of card games . The improvement is based on a card game model that is a structured representation of words and sentences from the rules of card games . In this paper, the card game model and the improved authoring system are introduced. The experimental evaluation of the improved system is also reported.

Learning by problem-posing is a promising way to learn arithmetic and mathematics. In this paper, we report the design and development of an interactive learning environment for problem-posing as sentence-integration in arithmetic word problems. In this research, we have paid a special attention to “reverse thinking” problems. To find the characteristics of the reverse thinking problems, we built a task model of problem-posing and design the learning environment based on the model. This leaning environment has been already developed and experimentally used in a class of fourth grade at an elementary school.

Several investigations have already suggested that question-posing is a promising learning as well as a problem-solving activity. In order to realize this learning method, however, how to give feedback for the posed questions becomes a big issue. In this paper, we describe an interactive learning environment that promotes learning by question-posing in English learning for beginners. The learning environment supports learners to pose question and answer sentences. The most important characteristics of the learning environment are automatic assessment of the posed questions and feedback based on the assessment. We call this assessment as agent-assessment. To realize these functions, we have used automatic question generation techniques from English sentences proposed in our previous researches. Preliminary evaluations and related works are also described in this paper.

Training in realistic drawing is a basic technique of art education. Even though a novice artist may feel something is wrong in his or her sketch, the novice often cannot explain what the errors are. It is typical for a drawing teacher to try to advise a student about errors during drawing training by using analogical explanation. Analogical explanation provides the learner with an image of a solid that reflects his or her errors. However, the significance or merit of this kind of explanation is not well understood, even though the importance of learning by error is well known in practical education conditions. This paper describes a technique for error awareness in pencil drawing. The technique involves the creation of a three-dimensional model that reflects the learner’s errors as acquired from a scanned digital image of the learner’s pencil drawing.

It is well known that an assignment report has positive effect for learning. However, a great deal of work is required to evaluate and give feedback to it, because it is difficult to reach agreement on grading standards among setters (instructors), graders (teaching assistants), and participants in the course (students). Therefore, in this research, we have been developing an evaluation support system of the assignment reports that is used by the teaching assistants. The system provides a report format that the students are able to follow to complete the report. The format allows the teaching assistants to insert their comments in a more visualized form that is easier to be understood by the students. This paper shows how the results generated from this correction support system are different from those generated by the conventional ways of report checking.

In this paper, we describe developing and evaluation of a learning environment for Solution-based Problem-Posing in the case of multi-digit subtraction. One of the ways to learn by problem-posing is to make learners pose problems which can be solved by a solution. We call the type of problem posing Solution-based Problem-osing. The Solution-based Problem-Posing helps the learners understand the solution. In the case of multi-digit subtraction, however, it is difficult to provide the solutions directly because the solutions in the multi-digit subtraction are complex for learners who are elementary school students when they learn them. We suggest a method they are required to solve problems to be conscious the solution of the problems and then pose problems based on the solution.

We have already proposed and implemented an automatic generation method of learning games. However, we also found that a few generated learning games were not useful for learning. Therefore, we propose detection method of the low efficient learning games and improvement method of the detected games in this paper. Automatic detection function has been implemented. We also report the results of experimental evaluation of the detection function and the improvement method of the detected games.

To make a new problem from the original one and to compare their solutions are promising activities to promote a learner to be aware of the structure of these problems. Especially for knowledge-rich problems like word problems in arithmetic, mathematics or physics, this awareness is very important to master the use of solution methods. To realize such exercises in physics, we have developed a computer-based learning environment that allows a learner to make a new problem by changing the original one and that diagnoses the problem change. Preliminary evaluation of the learning environment has also been reported.

In this paper, we have described Kit-Build method as an approach to realize automatic assessment of a concept map. In Kit-Build method, an ideal concept map (goal map) is prepared by an expert of a target domain at first, and parts are generated by decomposing the goal map. The parts are provided to a learner, and then the learner rebuilds a concept map (learner map) by connecting the parts. It would be easy to diagnose the maps by comparing the goal map and learner map because the same parts are used in them. We have already designed and developed a learning environment with the Kit-Build method. An experimental evaluation of the learning environment is also reported.

Error-based simulation (EBS) has been developed to generate phenomenon by using students' erroneous idea and also it is considered as a promising method to promote the students be aware of their errors. In this paper, we used the EBS in learning of “normal reaction force” in a junior high school, and then compared the EBS class (where students learned the concept with EBS) and the usual class (where students learned the concept as usual) by using the results of a pre-test, post-test, delayed post-test and an interview after the delayed post-test. The results suggest that EBS contributed to generalization and retention of the learned concept as for “normal reaction force”.

Because of the recent progress of inter-university communication system on the Internet or communication satellite system, lecture archiving system is becoming popular. Since the number of lectures is not small to record them manually by deploying technical staffs for each lecture, lectures are often recorded with a fixed video camera, which is always shooting wide-angle view including a projector screen or a blackboard. This kind of inflexible system might be acceptable when the main purpose of archiving is to record materials or lecture given by a lecturer. However, when the style of the lecture is rather based on communication between a lecturer and students, i.e., lectures that accepts questions, shooting not only the lecturer but also a student in questioning is mandatory to convey full circumstance of the lecture with nonverbal information. We describe a system that enables to track a person in questioning in lecture. By tracking a microphone with IR tag by an IR camera, the position of a person in questioning is tracked and he/she is shot by another visible ray camera synchronized with the IR camera.

In this paper, an error-awareness learning environment targeting the bugs of multi-digit subtraction is described. The authors first introduce the block-box model where the minuend is embodied as blocks and the subtrahend is embodied as boxes. In the block-box model, the process of subtraction is embodied as concrete manipulation of putting the blocks in the boxes. The bugs in multi-digit subtraction can be visualized as irregular manipulation of the blocks and boxes. They have developed a learning environment where a learner subtracts a multi-digit number described with blocks and boxes by direct-manipulation. The learner then can reconstruct an erroneous subtraction including a bug with irregular manipulation of the blocks and boxes. Through this reconstruction, they expect that a learner will increase his/her understanding about the procedure of subtraction. In this paper, an experimental evaluation of the learning environment is also described.

This paper presents the method of simulating the simultaneous equations which are over/under-constrained by the learner's misconception (partial constraint analysis: PCA). It can deal with arbitrary linear differential and/or algebraic equations with erroneous parts which preserve linearity. The procedure is as follows: (1) to generate the network which describes the constraints among the variables in equations (causal relation network: cn); (2) to abstract the subgraphs in which variables can be calculated (executed) independently of the rest (partial causal chains: pc); (3) to enumerate the combinations of pces which can be executed simultaneously; and (4) to execute the selected combination by supplying the driving data exogenously for simulating the behavior of the erroneous system. An example of application to a simple electric circuit is also presented.

This paper describes a computational method of complexity of a question for an adaptive question and answer function in an intelligent support system for English learning, and its evaluation. To realize adaptive question and answer, systems should generate questions depending on both educational intentions and learner's understanding state. For generating suitable questions for a learner automatically, systems must know the factors which influence difficulty of questions, and have the ability to calculate difficulty. Difficulty is composed of a learner dependent part and an independent part. The former is evaluated by referring to a student model. The latter is defined by enumerating factors which influence complexity of questions. We present a definition of complexity of questions along learners' answering process. We also describe experimentation of comparing the complexity of questions calculated by computer with the complexity evaluated by human.

A learner who can solve a problem cannot always understand the problem adequately. In order to deeply understand a problem, it is effective that he/she poses problems which can be solved by the solution method of the problem. Moreover after solving and posing problems, it is important for him/her to consider and indicate relations between the problems solved or posed by him/her. In order to let the learner think about the relations, the authors propose a support environment for learning by describing "problem map". The "problem map" is a kind of semantic network where a node is a problem and a link describes relations between two problems. The authors are developing a support environment for describing a problem map in arithmetic learning. The environment cannot only provide learners with workspace to describe problem map, but also feedback based on diagnosis of the described problem map. Additionally, the problem map can be used as a structured problem source for adaptive Web-based problem practice which is generated based on the problem map.

Learning games are promising to realize highly motivated learning. We categorize the methods to design computer-based learning game in the following three: (1) combination method, (2) fusion method, and (3) simple simulation method. Fusion method is a method to design the learning game such that the learning skills of this game are regarded as a learning activity. We have developed a learning game with this method targeting learning by problem posing of arithmetical word problems. In this paper, we explain the fusion method as a design method of a computer-based learning game. Then, the learning game of problem posing of arithmetical word problems is introduced as a design example of the fusion method. Cases of the use of the game at elementary school are also reported.

Modeling of the user’s interests is one of the most important issues to support a user to gather information. Although there are several promising methods to infer the interests from the user’s browsing behavior, they assume that the interests are consistent during the information gathering. However, in browsing which is one of the most popular ways to gather information, the user’s interests are often strongly dependent on the local context of the browsing. This paper describes a method to model the user’s shifting interests from the browsing history. An information filtering method using the model of the interests has been implemented. We call it “context-sensitive filtering.” The results of an experimental evaluation, by real users’ browsing for an encyclopedia in CDROM format, are also reported.

No abstract available.