Van-Doan Nguyen

A Community-Based Collaborative Filtering System Dealing with Sparsity Problem and Data Imperfections

In this paper, we develop a collaborative filtering system for not only tackling the sparsity problem by exploiting community context information but for also dealing with data imperfections by means of Dempster-Shafer theory. The experimental results show that the proposed system achieves better performance when comparing it with a similar system, CoFiDS.

Integrating with Social Network to Enhance Recommender System Based-on Dempster-Shafer Theory

In this paper, we developed a new collaborative filtering recommender system integrating with a social network that contains all users. In this system, user preferences and community preferences extracted from the social network are modeled as mass functions, and Dempster’s rule of combination is selected for fusing the preferences. Especially, with the community preferences, both the sparsity and cold-start problems are completely eliminated. So as to evaluate and demonstrate the advantage of the new system, we have conducted a range of experiments using Flixster data set.

Two-probabilities focused combination in recommender systems

In this paper, we propose a new method called 2-probabilities focused combination for combining information about user preferences on products or services in recommender systems based on Dempster-Shafer theory. Regarding this method, in focal sets of mass functions representing user preferences, focal elements with probabilities in top two highest ones are retained and the remaining focal elements are considered as noise and then transferred to the whole set element. To demonstrate the advantages of the new method, a baseline known as 2-points focused combination is selected for performance comparison in a range of experiments using Movielens and Flixster data sets. According to the results of experiments, the new method is more effective in accuracy of recommendations and comparable in computational time. Also, the new method is capable of overcoming the weakness of the baseline because of the ability to generate stable results. This paper proposes a new combination method for recommender systems based on Dempster-Shafer theory.The new method is capable of (1) handling combining highly conflicting mass functions, (2) improving computational time, and (3) overcoming the weakness of 2-points focused combination method.The new method is also experimentally tested on two different data sets.

Evidence Combination Focusing on Significant Focal Elements for Recommender Systems

In this paper, we develop a solution for evidence combination, called 2-probabilities focused combination, that concentrates on significant focal elements only. Firstly, in the focal set of each mass function, elements with their probabilities in top two highest probabilities are retained; others are considered as noise, which have been generated when assigning probabilities to the mass function and/or by related evidence combination tasks had already been done before, and eliminated. The probabilities of eliminated elements are added to the probability of the whole set element. The achieved mass functions are called 2-probabilities focused mass functions. Secondly, Dempster’s rule of combination is used to combine pieces of evidence represented as 2-probabilities focused mass functions. Finally, the combination result is transformed into the corresponding 2-probabilities focused mass function. Actually, the proposed solution can be employed as a useful tool for fusing pieces of evidence in recommender systems using soft ratings based on Dempster-Shafer theory; thus, we also present a way to integrate the proposed solution into these systems. Besides, the experimental results show that the performance of the proposed solution is more effective than a typically alternative solution called 2-points focused combination solution.

Learning structure-property relationship in crystalline materials: A study of lanthanide–transition metal alloys

We have developed a descriptor named Orbital Field Matrix (OFM) for representing material structures in datasets of multi-element materials. The descriptor is based on the information regarding atomic valence shell electrons and their coordination. In this work, we develop an extension of OFM called OFM1. We have shown that these descriptors are highly applicable in predicting the physical properties of materials and in providing insights on the materials space by mapping into a low embedded dimensional space. Our experiments with transition metal/lanthanide metal alloys show that the local magnetic moments and formation energies can be accurately reproduced using simple nearest-neighbor regression, thus confirming the relevance of our descriptors. Using kernel ridge regressions, we could accurately reproduce formation energies and local magnetic moments calculated based on first-principles, with mean absolute errors of 0.03 μB and 0.10 eV/atom, respectively. We show that meaningful low-dimensional representations can be extracted from the original descriptor using descriptive learning algorithms. Intuitive prehension on the materials space, qualitative evaluation on the similarities in local structures or crystalline materials, and inference in the designing of new materials by element substitution can be performed effectively based on these low-dimensional representations.

Using community preference for overcoming sparsity and cold-start problems in collaborative filtering system offering soft ratings

This paper proposes a new collaborative filtering recommender system offering soft ratings.In the system, community preferences are used for overcoming sparsity and cold-start problems.The new system is experimentally tested on Flixster data set. This paper introduces a new collaborative filtering recommender system that is capable of offering soft ratings as well as integrating with a social network containing all users. Offering soft ratings is known as a new methodology for modeling subjective, qualitative, and imperfect information about user preferences, as well as a more realistic and flexible means for users to express their preferences on products and services. Additionally, in the system, community preferences that are extracted from the social network are employed for overcoming sparsity and cold-start problems. In the experiment, the new system is tested using a data set culled from Flixster, a social network focused on movies. The experiments results show that this system is more effective than the selected baseline in terms of recommendation accuracy.

Noise-Averse Combination Method

This paper proposes a new method for combining information in recommender systems based on Dempster-Shafer theory. Within this method, focal elements whose probabilities are less than or equal to an infinitesimal threshold are considered as noise that may be caused by the process of fusing information, and then eliminated. Comparing with two baselines, known as 2-points focused and 2-probabilities focused combination methods, the new method is more effective because it does not eliminate focal elements whose probabilities are not very small. Moreover, the new method is tested in a wide range of experiments on MovieLens data set, and the experimental results indicate that the proposed method is much better than the baselines in accuracy of recommendations.

On Information Fusion in Recommender Systems Based on Dempster-Shafer Theory

In this paper, we address the problem of combining information in recommender systems (RSs) based on Dempster-Shafer theory (DST). We first discuss the characteristics of this problem, and then analyze six popular combination methods in the context of RSs. Based on the analysis, we propose two new mixed combination methods which can be considered as useful tools for fusing information in the systems. To evaluate the proposed methods, we integrate them into a typical RS based on DST, and then measure recommendation performances on MovieLens data set. The experimental results show that, comparing to the baselines, the new methods outperform with regards to DS-MAE and DS-Recall, and can be comparable in terms of DS-Precision and DS-F1.