Francisco Serradilla

Collaborative filtering adapted to recommender systems of e-learning

In the context of e-learning recommender systems, we propose that the users with greater knowledge (for example, those who have obtained better results in various tests) have greater weight in the calculation of the recommendations than the users with less knowledge. To achieve this objective, we have designed some new equations in the nucleus of the memory-based collaborative filtering, in such a way that the existent equations are extended to collect and process the information relative to the scores obtained by each user in a variable number of level tests.

Choice of metrics used in collaborative filtering and their impact on recommender systems

The capacity of recommender systems to make correct predictions is essentially determined by the quality and suitability of the collaborative filtering that implements them. The common memory-based metrics are Pearson correlation and cosine, however, their use is not always the most appropriate or sufficiently justified. In this paper, we analyze these two metrics together with the less common mean squared difference (MSD) to discover their advantages and drawbacks in very important aspects such as the impact when introducing different values of k-neighborhoods, minimization of the MAE error, capacity to carry out a sufficient number of predictions, percentage of correct and incorrect predictions and behavior when attempting to recommend the n-best items. The paper lists the results and practical conclusions that have been obtained after carrying out a comparative study of the metrics based on 135 experiments on the MovieLens database of 100,000 ratios.

Floating Car Data Augmentation Based on Infrastructure Sensors and Neural Networks

The development of new-generation intelligent vehicle technologies will lead to a better level of road safety and CO2 emission reductions. However, the weak point of all these systems is their need for comprehensive and reliable data. For traffic data acquisition, two sources are currently available: (1) infrastructure sensors and (2) floating vehicles. The former consists of a set of fixed point detectors installed in the roads, and the latter consists of the use of mobile probe vehicles as mobile sensors. However, both systems still have some deficiencies. The infrastructure sensors retrieve information from static points of the road, which are spaced, in some cases, kilometers apart. This means that the picture of the actual traffic situation is not a real one. This deficiency is corrected by floating cars, which retrieve dynamic information on the traffic situation. Unfortunately, the number of floating data vehicles currently available is too small and insufficient to give a complete picture of the road traffic. In this paper, we present a floating car data (FCD) augmentation system that combines information from floating data vehicles and infrastructure sensors, and that, by using neural networks, is capable of incrementing the amount of FCD with virtual information. This system has been implemented and tested on actual roads, and the results show little difference between the data supplied by the floating vehicles and the virtual vehicles.

Vehicle to Vehicle GeoNetworking using Wireless Sensor Networks

Vehicular communications will be the next quality step in the development of automotive technologies. Defining these communications is currently in the final step of development, the focus being on standardization and field tests of network devices and Advanced Driver Assistance Systems. However, some issues regarding vehicular communications that require a specific research effort are still open and represent a challenge if the technology is to be ultimately implemented and marketed. One of these challenges is to develop effective GeoNetworking in vehicular communications. This concept means that the Vehicle Ad-hoc NETwork (VANET) data package transmission is organized according to the topographical location of the different network nodes (vehicles), with the data flow being organized optimally so as to cover the surroundings of each vehicle. The core of this GeoNetworking system is the GeoRouting algorithm that supports the optimal routing of the data packages and reorganizes the network structure in accordance with the positions of the nodes. In this paper we present a novel GeoRouting algorithm for unicast communications, based on the evolution of the previous results of vehicular mesh networks using IEEE 802.15.4 Wireless Sensor Networks (WSN) technology. This algorithm has been designed, implemented and validated under controlled conditions and tested in real vehicles on real roads with free-flow traffic. The results suggest that the features of this routing algorithm can be inserted into any vehicular architecture to provide functioning GeoNetworking that will support a wide range of Advanced Driving Assistance System (ADAS) applications.

Multimodal Agents in Second Life and the New Agents of Virtual 3D Environments

The confluence of 3D virtual worlds with social networks imposes to software agents, in addition to his conversational functions, the same behaviors as those common to human-driven avatars. In this paper we explore the possibilities of the use of metabots (metaverse robots) in virtual 3D worlds and we introduce the concept of AvatarRank as a measure of the avatars popularity and the concept of the extended Turing test to assess the anthropomorphness of the metaverses elements.

Energy Consumption Estimation in Electric Vehicles Considering Driving Style

The limitation on the range of electric vehicles makes quite important to use an accurate energy consumption estimation tool. In general, estimations are based solely on the total distance, although it is known that the characteristics of the route and driving style influence significantly on energy consumption. In this paper, a tool that estimates the energy consumption of an electric vehicle in a city route taking into account such variables is shown, but without needing a deterministic knowledge of the characteristics of the vehicle or driving cycles. To do this, a neural network that takes as input data driving style and route variables is used. The validation results have been quite satisfactory to increase reliability in predicting consumption of the vehicle and enhance user confidence in the capabilities of electric vehicles.

Modeling spatial-temporal context information in virtual worlds

Currently, we use many definitions with diffuse boundaries: Web 2.0 (Social Networks), Web 3.0 (Semantic Web), Web 3D (Metaverses, Virtual Worlds, Mirror Worlds), Recommendation Systems, Augmented Reality, Geo-location... In this paper we explore the possibilities of the combined use of these concepts, we introduce the concept of VARD (Virtual Augmented Reality Device) and show interoperability between recommendation systems and Virtual Worlds. We have developed a Recommendation System which have two ways of interaction with the virtual world of Second Life in connection with context spatial-temporal information: an active recommendation system, called TESLAR, that interacts with avatars by a 2D HUD VARD object, and a passive and automatic recommendation system, called MarvinBot, that interacts with avatars by a Metabot VARD.

Cognitive Modeling for Navigation of Mobile Robots Using the Sensory Gradient Concept

In order to build models reflecting accurately the structure of the real world Artificial Intelligent-based systems have the difficult problem of reducing the continuous, extremely complex sensory information into a discrete and simple model with the optimal computational burden. The new concept of sensory gradient is introduced in this paper to obtain computational models of the sensory information available for an autonomous mobile robot. The main objective is to guarantee that these models are simple and at the same time powerful enough to allow the performance of complex navigation tasks. We use the sensory gradients module to identify situations and places that must be recorded (we call them Relevant Sensory Places or RSPs for short) and we build upon these RSPs a graph-based model of the universe that allows to develop navigation plans using plan-as-communication techniques. This novel approach is applied to the spatial reasoning problem as an aid to the autonomous navigation of mobile robots. Using a simulation environment the paper presents some empirical results which have been very encouraging. This novel approach has been successfully tested on a NOMAD-200 mobile robot platform.

Intravehicular, Short- and Long-Range Communication Information Fusion for Providing Safe Speed Warnings

Inappropriate speed is a relevant concurrent factor in many traffic accidents. Moreover, in recent years, traffic accidents numbers in Spain have fallen sharply, but this reduction has not been so significant on single carriageway roads. These infrastructures have less equipment than high-capacity roads, therefore measures to reduce accidents on them should be implemented in vehicles. This article describes the development and analysis of the impact on the driver of a warning system for the safe speed on each road section in terms of geometry, the presence of traffic jams, weather conditions, type of vehicle and actual driving conditions. This system is based on an application for smartphones and includes knowledge of the vehicle position via Ground Positioning System (GPS), access to intravehicular information from onboard sensors through the Controller Area Network (CAN) bus, vehicle data entry by the driver, access to roadside information (short-range communications) and access to a centralized server with information about the road in the current and following sections of the route (long-range communications). Using this information, the system calculates the safe speed, recommends the appropriate speed in advance in the following sections and provides warnings to the driver. Finally, data are sent from vehicles to a server to generate new information to disseminate to other users or to supervise drivers’ behaviour. Tests in a driving simulator have been used to define the system warnings and Human Machine Interface (HMI) and final tests have been performed on real roads in order to analyze the effect of the system on driver behavior.

Adaptive fuzzy knowledge‐based systems for control metabots' mobility on virtual environments

: The confluence of three-dimensional (3D) virtual worlds with social networks imposes on software agents, in addition to conversational functions, the same behaviours as those common to human-driven avatars. In this paper, we explore the possibilities of the use of metabots (metaverse robots) with motion capabilities in complex virtual 3D worlds and we put forward a learning model based on the techniques used in evolutionary computation for optimizing the fuzzy controllers which will subsequently be used by metabots for moving around a virtual environment.