1,720,977 research outputs found
Triadic Motifs in the Partitioned World Trade Web
No full textAbstractOne of the crucial aspects of the Internet of Things that influences the effectiveness of communication among devices is the communication model, for which no universal solution exists. The actual interaction pattern can in general be represented as a directed graph, whose nodes represent the “Things” and whose directed edges represent the sent messages. Frequent patterns can identify channels or infrastructures to be strengthened and can help in choosing the most suitable message routing schema or network protocol. In general, frequent patterns have been called motifs and overrepresented motifs have been recognized to be the low-level building blocks of networks and to be useful to explain many of their properties, playing a relevant role in determining their dynamic and evolution. In this paper triadic motifs are found first partitioning a network by strength of connections and then analyzing the partitions separately. The case study is the World Trade Web (WTW), that is the directed graph connecting world Countries with trade relationships, with the aim of finding its topological characterization in terms of motifs and isolating the key factors underlying its evolution. The WTW has been split based on the weights of the graph to highlight structural differences between the big players in terms of volumes of trade and the rest of the world. As test case, the period 2003-2010 has been analyzed, to show the structural effect of the economical crisis in the year 2007
Adjusted F-measure and kernel scaling for imbalanced data learning
No full textRare events are involved in many challenging real world classification problems, where the minority class is usually the most expensive to sample and to label. As a consequence, training data are often imbalanced, presenting an heavily skewed distribution of labels. Using conventional classification techniques produces biased results, as the classifier may easily show a very good performance on the over-represented class and a very poor performance on the under-represented class: the former dominates the learning process and tends to attract all predictions. Furthermore, the classical accuracy measure is misleading, as it assumes equal importance for the true positives and the true negatives. We propose a classification procedure based on Support Vector Machine able to effectively cope with data imbalance. Using a first step approximate solution and then a suitable kernel transformation, we enlarge asymmetrically space around the class boundary, compensating data skewness. We also propose an accuracy measure, named AGF, that properly accounts for the different misclassification costs of the two classes. Tests on real world data from a public repository show that the proposed approach outperforms its competitors
Generation of description metadata for video files
No full textAutomatic Metadata Generation in the context of e-learning standards is usually referred to algorithms able to process and annotate semi structured documents in plain text. As most of the information available on the web nowadays is unstructured and in the form of multimedia files, the need for more general approaches arises. We propose an automatic metadata generation procedure that allows to label specific unstructured data (video lectures) with metadata compliant to the Learning Object Metadata standard. After preprocessing, three different summarization algorithms are tested and used to obtain a synthetic description of video content, both in terms of Description and Title. Results show that, in the provided context, the obtained Description has a good agreement with the lesson abstract written by its author
Attributed Relational SIFT-Based Regions Graph for Art Painting Retrieval
No full textRecently, image retrieval and analysis algorithms have been extensively applied to art related domains. In this field, state-of-the-art approaches mainly focus on feature extraction with the aim of improving reliability of authentication, classification and retrieval of art paintings. In this paper we propose an effective modeling, based on a graph structure, and a retrieval strategy, based on a graph matching algorithm, for art paintings. The proposed approach has been tested on different datasets with high quality results allowing an user to run effective content-based queries on painting records
User click modeling on a learning management system
No full textClicking behavior is of prominent interest in many research fields. When accessing resources in a Learning Management Systems (LMS), clicks represent implicit feedbacks and carry precious information to improve content and layout, so to increase both the overall user experience (UX) and the resource effectiveness. As differences in age and cultural background are well known to affect the clicking behavior, the study of a homogeneous population allows to fully characterize it within a precisely delimited task. In the following, the pattern of access to learning resources of a group of graduated students involved in a specialized course is derived from log data, estimating its main behavioral stages, called orientation, evaluation and assimilation, and the transition rate from the first one to the next. Some statistics (average session time, total time of fruition and number of sessions) are also derived from the clicking distribution
Automatic generation of SCORM compliant metadata for portable document format files
No full textThe Shareable Content Object Reference Model (SCORM) is a widely adopted collection of specifications for web-based e-learning to which most Learning Management Systems adhere. While it allows reusability of content, it requires extensive, slow and expensive metadata annotation, and this fact prevents many content producers from properly creating and using Learning Objects. We propose an automatic metadata generation procedure that allows to label specific Learning Objects (scientific papers) with general metadata compliant to the SCORM. As some metadata are intrinsically unrelated to structure while others are strictly connected to structure, two different techniques were developed: one based on vocabularies and the other based on structural features. Results show that, in the provided context and for the "general" metadata category, the accuracy of annotations is comparable to that of a human expert
A Novel Graph-Based Fisher Kernel Method for Semi-supervised Learning
No full textGraph-based semi-supervised learning methods play a key role in machine learning applications, particularly when no parametric information or other prior knowledge is available. Given a graph whose nodes represent the points and the weighted edges the relations between them, the goal is to predict the values of all unlabeled nodes exploiting the information provided by both label and unlabeled nodes. In this paper, we propose a novel graph-based approach for semi-supervised binary classification. The algorithm extends the Fisher Subspace estimation approaches by adopting a kernel graph covariance measure. This similarity measure defines a relation between nodes generalizing both the shortest path and the commute time distance. This quantity is called the sum-over-paths covariance. Experiments on synthetic and real-world datasets highlight that the proposed algorithm achieves better results with respect to those obtained by state-of-the-art compet
Going Beyond Counting First Authors in Author Co-citation Analysis
Get PDFThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Tracking e-learning user sessions
No full textLearning Management Systems (LMS) are widely used to organise online learning materials within university courses. Most platforms do not include adequate features for access tracking and log analysis. An algorithm to generate session data and to compute synthetic statistics from these data is here proposed. The benefit is twofold: it allows teachers to evaluate usefulness of published resources and students to compare to others their own progresses. The case study on a course held within a research project at the University of Naples "Parthenope" showed encouraging results
- …
