Catálogo de publicaciones - libros

We address the task of improving convergence performance of PageRank computation. Based on a step-length calculation approach, we derive three methods, which respectively calculates its step-length so as to make the successive search directions orthogonal (orthogonal direction), minimize the error at the next iteration (minimum error) and make the successive search directions conjugate (conjugate direction). In our experiments using a real Web network, we show that the minimum error method is promising for this task.

-glycosylation is one of the main types of the mammalian protein glycosylation, which is serine or threonine specific, though any consensus sequence is still unknown. In this paper, a layered neural network and a support vector machine are used for the prediction of -glycosylation sites. Three types of encoding for a protein sequence within a fixed size window are used as the input to the network, that is, a sparse coding which distinguishes all 20 amino acid residues, 5-letter coding and hydropathy coding. In the neural network, one output unit gives the prediction whether a particular site of serine or threonine is glycosylated, while SVM classifies into the 2 classes. The performance is evaluated by the Matthews correlation coefficient. The preliminary results on the neural network show the better performance of the sparse and 5-letter codings compared with the hydropathy coding, while the improvement according to the window size is shown to be limited to a certain extent by SVM.

This paper proposes a method for sensitivity communication robots which infer their dialogist’s emotion. The method is based on the Bayesian approach: by using a Bayesian modeling for prosodic features. In this research, we focus the elements of emotion included in dialogist’s voice. Thus, as training datasets for learning Bayesian networks, we extract prosodic feature quantities from emotionally expressive voice data. Our method learns the dependence and its strength between dialogist’s utterance and his emotion, by building Bayesian networks. Bayesian information criterion, one of the information theoretical model selection method, is used in the building Bayesian networks. The paper finally proposes a reasoner to infer dialogist’s emotion by using a Bayesian network for prosodic features of the dialogist’s voice. The paper also reports some empirical reasoning performance.

We present a method for discovering nominally conditioned polynomials to fit multivariate data containing numeric and nominal variables using a four-layer perceptron having shared weights. A polynomial is accompanied with the nominal condition stating a subspace where the polynomial is applied. To get a succinct neural network, we focus on weight sharing, where a weight is allowed to have one of common weights. A near-zero common weight can be eliminated. Our method iteratively merges and splits common weights based on 2nd-order criteria, escaping from local optima. Moreover, our method selects the optimal number of hidden units based on cross-validation. The experiments showed that our method can restore the original sharing structure for an artificial data set, and discovers rather succinct rules for a real data set.

The purpose of this study is to develop a computer support system for educating personnel who are involved in care management. We wish to develop a system in which know-how information can be shared. We consider that visualizing and showing care plans drawn up by experts in various forms allows a beginner to see the differences between their plans and an expert plan. Sharing know-how information is possible by recording, accumulating, and giving titles to what has been noticed in comparing documents. This function can visualize the similarities among documents that interpreted the results of an assessment, and can flexibly change different viewpoints. In order to promote user awareness, we mapped a user’s new document into a two-dimensional document space, and confirmed that the results of this mapping were appropriate.

In information ages, a computer is our active intelligent collaborator to work and learn something, so that learner’s intelligent ability must be taken into consideration even in education of information technologies (IT). Aim of this paper is to propose quantitative evaluation of the intelligent ability in computer usage. The author discuses some elements in the intelligent ability (such as quickness, logics, accuracy, memory, discernment, patience, decision, and inference), and formulates quantitative evaluation method of the elements with computer games. A tutor of IT education will be able to apply result of the research to computer assisted education for supporting learner’s intelligent activities.

E-Learning gives the opportunity to students to learn at any time and any place. Over 250 students have enrolled in Shinshu University Graduate School on the Internet (SUGSI), from which almost 90% are working professionals. Working professionals are usually highly motivated when they enroll in the distance learning program. However, an important problem is that their motivation might decrease and learning could stagnate. Currently, we are implementing a supporting program which can rapidly find out learners’ decreased motivation. In this paper, we describe the distant student support system we are deploying and give some initial results on its evaluation.

Self-directed learning often finishes with incomplete knowledge. It is hard for learners to be aware of the knowledge incompleteness. The main issue addressed in this paper is how to enhance their awareness. Our approach to this issue is to enable learners to publish knowledge obtained from their learning process to review it. The most important point towards knowledge publishing is how to represent knowledge so that the learners can reflect on it from another perspective. However, such knowledge representation would be troublesome for the learners. This paper discusses how to promote learners’ knowledge publishing by automatically generating a hypertext with TOC (Table Of Contents) as representation of knowledge they have learned in their learning process. Publishing the hypertext enables the learners to gain not only their own reviews but also the reviews from the peers.

One of today’s hot topics in the field of education is the learning support system. With the progress of networks and multimedia technologies, various types of web-based training (WBT) systems are being developed for distance- and self-learning. Most of the current learning support systems synchronously reproduce lecture resources such as videos, slides, and digital-ink notes written by the teacher. However, from the perspective of support for student learning, these systems provide only keyword retrieval. This paper describes a more efficient learning support system that we developed by introducing lecture information and student arousal levels extracted from biological data. We also demonstrate the effectiveness of the proposed system through a preliminary experiment.

We have been developing a system , a computer-mediated learning system that employs digital pen to realize paper-centric augmented classroom. Although the approach was sophisticated, it restricted feedback effects which potentially improve learning. To maximize the feedback effects of the system, we present a hybrid approach of augmented classroom environment. We classified the type of feedback loops in terms of the hybrid approach and improved the system to accomplish the functions by adding (1) note browsing interface for handheld, (2) worksheet editor for teachers, (3) handwriting character recognition engine for versatile use, and (4) HTTP embedded server function for flexible reference of collective results and notes. We also conducted a feasibility study and investigated the effectiveness of the immediate feedback under the hybrid approach.