María N. Moreno

A hybrid recommendation approach for a tourism system

Many current e-commerce systems provide personalization when their content is shown to users. In this sense, recommender systems make personalized suggestions and provide information of items available in the system. Nowadays, there is a vast amount of methods, including data mining techniques that can be employed for personalization in recommender systems. However, these methods are still quite vulnerable to some limitations and shortcomings related to recommender environment. In order to deal with some of them, in this work we implement a recommendation methodology in a recommender system for tourism, where classification based on association is applied. Classification based on association methods, also named associative classification methods, consist of an alternative data mining technique, which combines concepts from classification and association in order to allow association rules to be employed in a prediction context. The proposed methodology was evaluated in some case studies, where we could verify that it is able to shorten limitations presented in recommender systems and to enhance recommendation quality.

Making use of associative classifiers in order to alleviate typical drawbacks in recommender systems

Nowadays, there is a constant need for personalization in e-commerce systems. Recommender systems make suggestions and provide information about items available, however, many recommender techniques are still vulnerable to some shortcomings. In this work, we analyze how methods employed in these systems are affected by some typical drawbacks. Hence, we conduct a case study using data gathered from real recommender systems in order to investigate what machine learning methods can alleviate such drawbacks. Due to some especial features inherited by associative classifiers, we give a particular attention to this category of methods to test their capability of dealing with typical drawbacks.

A fuzzy associative classification approach for recommender systems

Despite the existence of dierent methods, including data mining techniques, available to be used in recommender systems, such systems still contain numerous limitations. They are in a constant need for personalization in order to make effective suggestions and to provide valuable information of items available. A way to reach such personalization is by means of an alternative data mining technique called classification based on association, which uses association rules in a prediction perspective. In this work we propose a hybrid methodology for recommender systems, which uses collaborative altering and content-based approaches in a joint method taking advantage from the strengths of both approaches. Moreover, we also employ fuzzy logic to enhance recommendations quality and eectiveness. In order to analyze the behavior of the techniques used in our methodology, we accomplished a case study using real data gathered from two recommender systems. Results revealed that such techniques can be applied eectively in recommender systems, minimizing the eects of typical drawbacks they present.

Determination of macroscopic thermodynamic ionization constants under conditions of variable ionic strength by an optimization algorithm

Abstract The thermodynamic macroscopic ionization constants of cysteine, penicillamine, 6-amino penicillanic acid and 7-amino cephalosporanic acid were determined at different temperatures by a dynamic method employing potentiometric data obtained at ionic strengths that vary continuously throughout the titration. This method gives thermodynamic pK-values in a single experiment. A computer program has been developed for the titrations. It makes use of an algorithm based on controlled descent.

Computation in kinetics, II. Simultaneous regression estimation of rate constants and initial concentrations

The general regression program KILET was used for simultaneous estimation of rate constants and initial concentrations. The method can be used either to improve the quality of fit calculating the possible analytical errors of some initial concentrations of a kinetic experiment or for analytical purposes, i.e. for the determination of analytes, while any side reactions can be taken into account.AbstractОбщую регрессионную программу KILET использовали для одновременного определения констант скоростей и нача-льных концентраций. Метод может быть использован либо для повышения качества совмещения с расчетами возмож-ных аналитических погрешностей некоторых начальных кон-центраций в кинетических экспериментах, либо для ана-литических целей, а именно для определения аналитов, в то время как все побочные реакции приняты во внимание.

Using Association Analysis of Web Data in Recommender Systems

The numerous web sites existing nowadays make available more information than a user can manage. Thus, an essential requirement of current web applications is to provide users with instruments for personalized selective retrieval of web information. In this paper, a procedure for making personalized recommendations is proposed. The method is based on building a predictive model from an association model of Web data. It uses a set of association rules generated by a data mining algorithm that discovers knowledge in an incremental way. These rules provide models with relevant patterns that minimize the recommendation errors.

HyCo – An Authoring Tool to Create Semantic Learning Objects for Web-Based E-learning Systems

In this article we introduce HyCo (Hypertext Composer), an authoring tool devoted to create semantic learning objects. This authoring tool uses learning technology standards or specifications to save these semantic objects, which will be delivered in Web e-learning environments as encapsulated packages in order to ensure their reusability, interoperability, durability and accessibility. These learning objects are closed to the Semantic Web field because they combine hypermedia and semantic capabilities. Our research work is directed to use these semantic learning objects in order to define learning domains for an Adaptive Learning Environment. The aim of this system is to provide an e-learning environment where teachers have tools to create didactic materials and students carry out their knowledge acquisition through the most suitable adaptive learning technique giving the student’s characteristics, the learning activities provided, and the learning objects’ features.

Information-Theoretic Measures for Meta-learning

Information-theoretic measures are suitable to characterize datasets with discrete attributes (or continuous which can be transformed). They can find information that can be decisive in order to analyze the behavior of different learning algorithms with specific datasets. The objective of the work presented in this paper is to study by means of three similar datasets from UCI Repository Machine Learning, the possible reasons for which breast-cancer-wisconsin dataset, in comparison with other 20 datasets, showed in a previous research that Stacking by Meta-Decision Trees (MDT) was significant better than all other multiclassifier models, including Stacking by Multi-Response Linear Regression (MLR). In our experiments the proportion of missing values, among other significant changes in different measure values, provided evidences about the possible origin of the different behaviors presented by these multiclassifier schemes depending on data characteristics.

Consecutive reactions in chemical kinetics: A method for the treatment of experimental data

A method is proposed for the treatment of data from a kinetic system of first order consecutive reactions by using an optimization procedure which converges to the correct solution of the two mathematically possible solutions and which does not need initial estimates of the parameters close to those to be determined.AbstractПредложен метод обработки данных кинетической системы последовательных реакций первого порядка, используя процедуру оптимализации, конвергирующую к правильному решению из двух математически возможных решений и не нуждающуюся в исходных приближениях параметров, близких к определяемым.

Multi-label Classification for Recommender Systems

Multi-label classification groups a set of supervised learning methods producing models capable of classifying examples in more than one class. These methods have been applied in diverse fields; however, the field of recommender systems has been hardly explored. In this work, books’ recommendation data are used to evaluate the behavior of the main multi-label classification methods in this application domain. The experiments carried out demonstrated their suitability to provide reliable recommendations and to avoid the grey sheep problem.