M. De Santo

A large database of graphs and its use for benchmarking graph isomorphism algorithms

Despite of the fact that graph-based methods are gaining more and more popularity in different scientific areas, it has to be considered that the choice of an appropriate algorithm for a given application is still the most crucial task.The lack of a large database of graphs makes the task of comparing the performance of different graph matching algorithms difficult, and often the selection of an algorithm is made on the basis of a few experimental results available.In this paper we present an experimental comparative evaluation of the performance of four graph matching algorithms. In order to perform this comparison, we have built and made available a large database of graphs, which is also described in detail in this article.

Evaluating on-line learning platforms: a case study

Our information-oriented society shows an increasing exigency of life-long learning. In such framework, online learning is becoming an important tool to allow the flexibility and quality requested by such a kind of learning process. In the recent past, a great number of on-line platforms have been introduced on the market showing different characteristics and services. A series of features should be taken into account when one evaluates e-learning platforms, starting from the function and usability of the overall learning system in the context of the human, social and cultural organization within which it is to be used. Obviously, the analysis of the features of a system is not sufficient: it is also important to understand how they are integrated to facilitate learning and training and what principles are applied to guide the way the system is used. To evaluate them both pedagogical and technological aspects must be carefully evaluated. This paper proposes a model for describing and characterizing on-line learning platform component. The model is then used to evaluate the most known existing commercial platforms.

Uncertainty characterization in image based measurements: a preliminary discussion

There are a large number of application fields where measurements deriving from digital images can assume a great relevance. Nevertheless, to make a profitable use of such measurements, it is indispensable to achieve a complete and quantitative control on uncertainties that real systems introduce along the chain of steps going from real-world objects to the results of the measurement process. This paper deals with this nontrivial task and, in particular, with the analytical expression of uncertainty characterizing the results of image processing software. At first, a simplified model of uncertainty of digital images is derived and experimentally tested; then the Canny edge detector output uncertainty is analytically expressed and verified both in artificial and real-world images.

Work in progress - virtual lab for electronic engineering curricula

During last years the interest on distance learning techniques has grown steadily as far as the use of electronic instruments in experimentation is concerned. Due to the higher and higher number of students accessing the university educational structures, the cost of laboratories for didactical electronic applications is going to be very high. As a consequence, a number of software tools and environments have been developed to help users to share distributed laboratory resources and realize virtual experiments. Nevertheless, further solutions have to be explored when students must be trained and experienced in the instrumentation programming. In this paper, we exploit modern software technologies to design and implement a distributed architecture for virtual labs allowing the approach previously described. Services integrated in this architecture aim to support students both to keep contact with real instruments both to remotely program instrumentation. This distance learning methodology is discussed and some reports from students experience with the system are showed.

Adaptive hypermedia system in education: A user model and tracking strategy proposal

New technologies are quickly changing the traditional educational approaches and systems. In fact the integration of new technologies in the field of education offers new challenges and opportunities in distance learning, lifelong learning and e-learning in general. On the other hand e-Learning is an interesting application area for adaptive hypermedia system (AEH). Through the use of user models intelligent tutoring and students tracking techniques an AEH can recognize individual users and their needs adapting in an effective way the student learning path. This paper attempts to enhance student learning by addressing different learning styles by the use of the Adaptive Hypermedia System approach. In particular it addresses to define a standardized and simple user model and some tracking parameters and techniques in order to update the proposed learning path. The proposed system collects information on the user learning style by the use of various well known in literature test We propose a mapping of this acquired information in some parameters defined in standard metadata user description like IMS LIP. The final goal is a student learning style representation by the use of a quantitative model So we can develop a tracking strategy through the observation of the main model parameters. A similar approach could be used for the description of learning objects in order to provide the introduction of well defined metrics for the dynamic tailoring of the learning path to the students learning style and needs.

Joint audio-video processing of MPEG encoded sequences

The current research efforts in the field of video parsing and analysis are focused on the use of pictorial information, while neglecting an important supplementary source of content information such as the embedded audio or soundtrack. In contrast, we address the issue of scene change detection with the use of video and audio information. We also discuss how joint exploitation of audio and video can be thoroughly performed on MPEG encoded video sequences. First experimental results are presented and discussed.

A Multi-expert Approach for Shot Classification in News Videos

In this paper we propose a multi-expert approach for anchor shot detection in news videos. The proposed Multi-Expert System (MES) combines three algorithms selected among those presented in the literature that are model-free and do not require a specific training phase. In order to assess the performance of the proposed MES, we built up a large database, significantly wider than those ones typically used in the field. Reported experimental results show not only that proposed system performances are better than the ones of original experts but also the smart use made by the MES of strengths and weaknesses of each expert.

Experimental characterization of synchronization protocols for wireless networks

Nowadays more and more measurement applications where sensor nodes cooperate to realize a data fusion process are entrusted to wireless sensor networks (WSN). Data fusion to be accurate requires that nodes be referred to a common clock and among themselves. In this scenario node synchronization got a crucial research topic and even sophisticated synchronization algorithms can be easily found in literature. Not a few others show to be especially designed for low cost, low memory and low power architectures. Their performance is usually evaluated either through analytical considerations or in simulation environments, but in some cases where ad hoc radio systems have been adopted. This paper makes a first step toward an experimental characterization of synchronization protocols that hold into account also influence factors which arise when commercial wireless communication busses as WiFi, BlueTooth (BT), ZigBee are adopted.

Unsupervised news video segmentation by combined audio-video analysis

Segmenting news video into stories is among key issues for achieving efficient treatment of news-based digital libraries. In this paper we present a novel unsupervised algorithm that combines audio and video information for automatic partitioning news videos into stories. The proposed algorithm is based on the detection of anchor shots within the video. In particular, a set of audio/video templates of anchorperson shots is first extracted in an unsupervised way, then shots are classified by comparing them to the templates using both video and audio similarity. Finally, a story is obtained by linking each anchor shot with all successive shots until another anchor shot, or the end of the news video, occurs. Audio similarity is evaluated by means of a new index and helps to achieve better performance in anchor shot detection than pure video approach. The method has been tested on a wide database and compared with other state-of-the-art algorithms, demonstrating its effectiveness with respect to them.

An agent based Internet infrastructure for learning commerce

Among the incredible amount of applications brought on the wave by the diffusion of the Internet, the possibility of an electronic commercialization of learning assumes a relevant role. Thousands of people potentially acting as learning material sellers, buyers and brokers are present on the Net asking for an efficient and effective infrastructure allowing them to play their role in this new market. Motivations are present and mature to investigate how to build such an infrastructure. We exploit agents and their mobility feature to develop a system for the electronic commerce of learning materials available on the Web. We tackle the aspects related to the negotiation, sometimes needed, to conclude a transaction and the aspects related to the best ways to save and classify the materials, through the learning object metadata approach and the related standardization efforts. A prototype architecture and its implementation is presented and discussed and directions for future studies are set.Among the incredible amount of applications brought on the wave by the diffusion of the Internet, the possibility of an electronic commercialization of learning assumes a relevant role. Thousands of people potentially acting as learning material sellers, buyers and brokers are present on the Net asking for an efficient and effective infrastructure allowing them to play their role in this new market. Motivations are present and mature to investigate how to build such an infrastructure. We exploit agents and their mobility feature to develop a system for the electronic commerce of learning materials available on the Web. We tackle the aspects related to the negotiation, sometimes needed, to conclude a transaction and the aspects related to the best ways to save and classify the materials, through the learning object metadata approach and the related standardization efforts. A prototype architecture and its implementation is presented and discussed and directions for future studies are set.