Efficient energy-based embedding models for link prediction in knowledge graphs

We focus on the problem of link prediction in Knowledge Graphs, with the goal of discovering new facts. To this purpose, Energy-Based Models for Knowledge Graphs that embed entities and relations in continuous vector spaces have been largely used. The main limitation in their applicability lies in the parameter learning phase, which may require a large amount of time for converging to optimal solutions. In this article, we first propose an unified view on different Energy-Based Embedding Models. Hence, for improving the model training phase, we propose the adoption of adaptive learning rates. We show that, by adopting adaptive learning rates during training, we can improve the efficiency of the parameter learning process by an order of magnitude, while leading to more accurate link prediction models in a significantly lower number of iterations. We extensively evaluate the proposed learning procedure on a variety of new models: our result show a significant improvement over state-of-the-art link prediction methods on two large Knowledge Graphs, namely WordNet and Freebase.     

Development of an auxiliary system for the execution of vascular catheter interventions with a reduced radiological risk; system description and first experimental results

Vascular catheterization is a common procedure in clinical medicine. It is normally performed by a specialist using an X-ray fluoroscopic guide and contrast-media. In the present paper, an image-guided navigation system which indicates a path providing guidance to the desired target inside the vascular tree is described with the aim of reducing the exposure of personnel and patients to X-rays during the catheterization procedure. A 3D model of the patient vascular tree, reconstructed with data collected by an angiography before starting the intervention, is used as a guide map instead of fluoroscopic scans. An accurate spatial correspondence between the body of the patient and the 3D reconstructed vascular model is established and, by means of a position indicator installed over the catheter tip, the real-time position/orientation of the tip is indicated correctly. This paper describes the system and the operational procedures necessary to use the proposed method efficiently during a catheter intervention. Preliminary experimental results on a phantom are also reported.     

Strain sensing fabric for hand posture and gesture monitoring.

In this paper, we report on a new technology used to implement strain sensors to be integrated in usual garments. A particular conductive mixture based on commercial products is realized and directly spread over a piece of fabric, which shows, after the treatment, piezoresistive properties, i.e., a change in resistance when it is strained. This property is exploited to realize sensorized garments such as gloves, leotards, and seat covers capable of reconstructing and monitoring body shape, posture, and gesture. In general, this technology is a good candidate for adherent wearable systems with excellent mechanical coupling with body surface. Here, we mainly focused on a sensorized glove able to detect posture and movements of the fingers. It could be used in several fields of application. We report on experimental results of a sensorized glove used as movements recorder for rehabilitation therapies and medicine. Furthermore, we describe a dedicated methodology used to read the output sensors which allowed to avoid using metallic wires for the connections. The price to be paid for all these advantages is a nonlinear electric response of the fabric sensor and a too long settling time, that in principle, make these sensors not suitable for real-time applications. Here we propose a hardware and computational solution to overcome this limitation.     

Analysis of Residual Stresses and Dislocation Density of AA6082 Butt Welds Produced by Friction Sir Welding

Metallurgical and Materials Transactions A - The Friction Stir Welding (FSW) method was employed to join AA6082 sheets. The welds were produced with different tool traverse speed (200 and...     

AN EMPIRICAL EVALUATION OF PARAMETER SENSITIVITY TO CHOICE SET DEFINITION IN SHOPPING DESTINATION CHOICE MODELS

ABSTRACT This paper empirically examines parameter sensitivity to choice set specification in the context of shopping destination choice, using supermarket choice data from Gainesville, Horida. We estimate parameters of the widely applied multinomial logit (MNL) discrete choice mode) multiple times. Each estimation uses, for all observations, a single randomly selected subset of the universal choice set. The distribution of parameter estimates is examined for specific market segments and choice subset sizes. The results indicate that the parameters of the model can be quite sensitive to the selection of the choice set used in the calibration. However, this sensitivity is not even across all parameters and there are some interesting variations. Distance deterrence and chain image parameters, for example, exhibit much more stability than parameters for store size and store competition. Li addition, model parameters show encouraging stability with relatively small choice sets of seven to ten stores.     

User profiling and virtual agents: a case study on e-commerce services

The main contribution of this work is the design of an application framework based on both conversational agents and user profiling technologies for the development of e-commerce services. User profiles are exploited by conversational agents to help customers in retrieving potentially interesting products from a catalogue. Three techniques were used for collecting data for a usability test: eye-movement tracking, questionnaire, and recording the user-system dialogue. The main outcomes of the experimental sessions are: (1) the dialogue capabilities of the agent facilitate the interaction between the user and the e-commerce site; and, (2) user profiles improve the retrieval capabilities of the agent. Finally, some limitations of the user profiling techniques adopted in the framework are discussed and a more sophisticated content-based profiling technique is proposed.