Relevance-based language modelling for recommender systems

Relevance-Based Language Models, commonly known as Relevance Models, are successful approaches to explicitly introduce the concept of relevance in the statistical Language Modelling framework of Information Retrieval. These models achieve state-of-the-art retrieval performance in the pseudo relevance feedback task. On the other hand, the field of recommender systems is a fertile research area where users are provided with personalised recommendations in several applications. In this paper, we propose an adaptation of the Relevance Modelling framework to effectively suggest recommendations to a user. We also propose a probabilistic clustering technique to perform the neighbour selection process as a way to achieve a better approximation of the set of relevant items in the pseudo relevance feedback process. These techniques, although well known in the Information Retrieval field, have not been applied yet to recommender systems, and, as the empirical evaluation results show, both proposals outperform individually several baseline methods. Furthermore, by combining both approaches even larger effectiveness improvements are achieved.     

Determining the minimum number of passive stretches necessary to alter musculotendinous stiffness

Abstract

In this study, we examined the minimum number of constant-torque passive stretches necessary to reduce musculotendinous stiffness. Thirteen healthy individuals (mean age 22 years, s = 3; stature 1.67 m, s = 0.1; mass 66 kg, s = 13 kg) volunteered to participate in the investigation and underwent four 30-s constant-torque passive stretches of the plantar flexor muscles. Musculotendinous stiffness was examined from the angle–torque curves generated prior to the passive stretches, at the beginning of each 30-s stretch, and immediately following the four 30-s passive stretches. The results indicated that musculotendinous stiffness of the plantar flexors was reduced following two 30-s constant-torque passive stretches (P < 0.05) compared with the pre- musculotendinous stiffness assessment. Musculotendinous stiffness remained depressed following the third and fourth stretches, but did not decrease further. These findings suggest that two 30-s bouts of constant-torque passive stretching may be necessary to cause a significant decrease in musculotendinous stiffness of the plantar flexor muscles.     

Sliding mode optimal control for linear systems

This paper addresses the optimal control problem for a linear system with respect to a Bolza–Meyer criterion with a non-quadratic non-integral term. The optimal solution is obtained as a sliding mode control, whereas the conventional linear feedback control fails to provide a causal solution. Performance of the obtained optimal controller is verified in the illustrative example against the conventional LQ regulator that is optimal for the quadratic Bolza–Meyer criterion. The simulation results confirm an advantage in favor of the designed sliding mode control.     

Some numerical tests for an alternative approach to optimal feedback control

We test a recently proposed approach to optimal feedback control of nonlinear systems leading to an iterative descending strategy [24]. We start by discussing the numerical implementation of this strategy, and propose a number of improvements that can speed up the computation process by up to two orders of magnitude. The resulting algorithm is then applied to a series of test problems of increasing complexity. Results seem to show that this can be a promising strategy to bear in mind for more realistic situations.     

Comparative analysis of academic grades in compulsory secondary education in Spain using statistical techniques

The present study, based on the construct comparability approach, performs a comparative analysis of general points average for seven courses, using exploratory factor analysis (EFA) and the Partial Credit model (PCM) with a sample of 1398 student subjects (M = 12.5, SD = 0.67) from 8 schools in the province of Alicante (Spain). EFA confirmed a one-factor model which explains 74.44% of the variance. Cronbach’s alpha value for this factor was .94. The PCM supported the one-factor model, and an optimal fit was achieved in all of the courses. The analysis of differential item functioning showed no significant differences in any course. Equitable distribution was observed in the evolution of the difficulty indices along the measurement scale for each course. This type of analysis confirms the measurement of a single latent construct in the different topics analysed, despite addressing various theoretical and procedural contents.     

Some challenges for educational research in Latin America

Founding director of the Centro de Estudios Educativos (Centre for Educational Studies) in Mexico City and of the Revista Latinoamericana de Estudios Educativos(Latin American Journal of Educational Studies). He was in charge of the National Programme of Educational Research of the National Council for Science and Technology. He now carries out research at the Regional Centre for Adult Education and Functional Literacy for Latin America. His recent publications include: La política de alfabetización de seis países latinoamericanos, Participatory Research: A New Research Paradigm?and Elementos para una propuesta orientada a elevar la calidad: La enseñanza tutorial.     

Investigating the factors that diminish the barriers to university-industry collaboration

Although the literature on university-industry links has begun to uncover the reasons for, and types of, collaboration between universities and businesses, it offers relatively little explanation of ways to reduce the barriers in these collaborations. This paper seeks to unpack the nature of the obstacles to collaborations between universities and industry, exploring influence of different mechanisms in lowering barriers related to the orientation of universities and to the transactions involved in working with university partners. Drawing on a large-scale survey and public records, this paper explores the effects of collaboration experience, breadth of interaction, and inter-organizational trust on lowering different types of barriers. The analysis shows that prior experience of collaborative research lowers orientation-related barriers and that greater levels of trust reduce both types of barriers studied. It also indicates that breadth of interaction diminishes the orientation-related, but increases transaction-related barriers. The paper explores the implications of these findings for policies aimed at facilitating university-industry collaboration.     

Modelling mathematical argumentation: the importance of qualification

In recent years several mathematics education researchers have attempted to analyse students’ arguments using a restricted form of Toulmin’s [The Uses of Argument, Cambridge University Press, UK, 1958] argumentation scheme. In this paper we report data from task-based interviews conducted with highly talented postgraduate mathematics students, and argue that a superior categorisation of genuine mathematical argumentation is provided by the use of Toulmin’s full scheme. In particular, we suggest that modal qualifiers play an important and previously unrecognised role in mathematical argumentation, and that one of the goals of instruction should be to develop students’ abilities to appropriately match up warrant-types with modal qualifiers.     

Building user profiles based on sequences for content and collaborative filtering

Modeling user profiles is a necessary step for most information filtering systems – such as recommender systems – to provide personalized recommendations. However, most of them work with users or items as vectors, by applying different types of mathematical operations between them and neglecting sequential or content-based information. Hence, in this paper we study how to propose an adaptive mechanism to obtain user sequences using different sources of information, allowing the generation of hybrid recommendations as a seamless, transparent technique from the system viewpoint. As a proof of concept, we develop the Longest Common Subsequence (LCS) algorithm as a similarity metric to compare the user sequences, where, in the process of adapting this algorithm to recommendation, we include different parameters to control the efficiency by reducing the information used in the algorithm (preference filter), to decide when a neighbor is considered useful enough to be included in the process (confidence filter), to identify whether two interactions are equivalent (δ-matching threshold), and to normalize the length of the LCS in a bounded interval (normalization functions). These parameters can be extended to work with any type of sequential algorithm.We evaluate our approach with several state-of-the-art recommendation algorithms using different evaluation metrics measuring the accuracy, diversity, and novelty of the recommendations, and analyze the impact of the proposed parameters. We have found that our approach offers a competitive performance, outperforming content, collaborative, and hybrid baselines, and producing positive results when either content- or rating-based information is exploited.