Kleanthi Lakiotaki

UTA-Rec: a recommender system based on multiple criteria analysis

UTARec, a Recommender System that incorporates Multiple Criteria Analysis methodologies is presented. The systems performance and capability of addressing certain shortfalls of existing Recommender Systems is demonstrated in the case of movie recommendations. UTARecs accuracy is measured in terms of Kendalls tau and ROC curve analysis and is also compared to a Multiple Rating Collaborative Filtering (MRCF) approach. The results indicate that the proposed Multiple Criteria Analysis methodology can certainly improve the recommendation process by producing highly accurate results, from a user oriented perspective.

Applications of MCDA in Marketing and e-Commerce

This chapter emphasizes on the major components under which MCDA applications in marketing and e-commerce have been developed and describes characteristic examples of research works that apply MCDA methodologies in marketing and e-commerce. The chapter is divided into two main sections separating the MCDA applications in the marketing discipline from those that appear in the ecommerce field. In each section fundamental notions of marketing and e-commerce are discussed accordingly and some characteristic examples of research works are analytically mentioned. The aim of this work is to endow candidate researchers that are interested in applyingMCDA methodologies in marketing and e-commerce with adequate background information to further develop their scopes and ideas.

Towards personalized medical document classification by leveraging UMLS semantic network

The overwhelmed amount of medical information available in the research literature, makes the use of automated information classification methods essential for both medical experts and novice users. This paper presents a method for classifying medical documents into documents for medical professionals (experts) and non-professionals (consumers), by representing them as term vectors and applying Multiple Criteria Decision Analysis (MCDA) tools to leverage this information. The results show that when medical documents are represented by terms extracted from AMTEx, a medical document indexing method, specifically designed for the automatic indexing of documents in large medical collections, such as MEDLINE, better classification performance is achieved, compared to MetaMap Transfer, the automatic mapping of biomedical documents to UMLS term concepts developed by U.S. National Library of Medicine, or the MeSH method, under which documents are indexed by human experts.

Analysing user behaviour in recommender systems

Nowadays, recommender systems are considered to be a valuable tool for internet marketing. Multi-criteria user modelling methodologies have been successfully applied to increase recommender systems accuracy. However, modelling user behaviour can be hard and often misleading when only the overall preference rate is considered. Various multi-criteria recommendation algorithms have been proposed that try to achieve high recommendation scores, but the gap from research ideas to real life applications remain large. Hence, studies concerning the understanding and interpretation of theoretical results together with direct application in real user data will improve and establish multi-criteria user profiling techniques as an important tool for recommender systems. In this direction, we analyse movie user profiles as a result of a multi-criteria recommendation methodology, applied to real user data, in order to reveal any hidden aspect of user behaviour that would eventually improve current system’s performance.

Students’ behaviour in peer assessment: a multi-criteria clustering approach

Peer assessment (PA), as formative procedure, enhances learning by providing students with the opportunity to peer assess each others work. However, since students exhibit different value systems (abilities, experiences, attitudes, cognitive styles, etc.) we propose a diagnostic procedure, which can be applied at the beginning of a course, in order to infer the most prevailing attitude among students. For this purpose, the proposed methodological framework, based on a multi-criteria clustering approach, identifies different assessment behaviours, in order to adopt the most common student assessment policy. To demonstrate the proposed method, an example is presented where two different PA data-sets are examined. The results clearly indicate that students exhibit different PA policies, nevertheless, a densely populated group is formed, the value system of which can be adopted thereafter.

BioDataome: a collection of uniformly preprocessed and automatically annotated datasets for data-driven biology

Abstract Biotechnology revolution generates a plethora of omics data with an exponential growth pace. Therefore, biological data mining demands automatic, ‘high quality’ curation efforts to organize biomedical knowledge into online databases. BioDataome is a database of uniformly preprocessed and disease-annotated omics data with the aim to promote and accelerate the reuse of public data. We followed the same preprocessing pipeline for each biological mart (microarray gene expression, RNA-Seq gene expression and DNA methylation) to produce ready for downstream analysis datasets and automatically annotated them with disease-ontology terms. We also designate datasets that share common samples and automatically discover control samples in case-control studies. Currently, BioDataome includes ∼5600 datasets, ∼260 000 samples spanning ∼500 diseases and can be easily used in large-scale massive experiments and meta-analysis. All datasets are publicly available for querying and downloading via BioDataome web application. We demonstrate BioDataome’s utility by presenting exploratory data analysis examples. We have also developed BioDataome R package found in: https://github.com/mensxmachina/BioDataome/. Database URL: http://dataome.mensxmachina.org/

Recommending medical documents by user profile

The overwhelmed amount of medical information available online, makes the use of automated recommendation methods essential for identifying relevant information according to user profile needs. This paper presents a method to address the problem of medical document classification into documents for medical professionals (experts) and non-professionals (consumers). Documents are represented by terms extracted from AMTEx, a medical document indexing method, specifically designed for the automatic indexing of documents in large medical collections, such as MEDLINE, and then mapped to the UMLS Semantic Network (SN) categories. Multiple Criteria Decision Analysis (MCDA) tools are applied to calculate the membership of each SN category to the document classification. Several factors such as the classification nature of the problem and the incorporation of common readability formulas are also examined.

A probabilistic choice model for the product line design problem

Designing optimal product lines is essential for any firm to stay competitive. Whereas a large number of optimization algorithms have been applied for solving the problem, most of them adopt the first choice rule to simulate the consumers choice process. Researchers avoid using probabilistic choice models, such as the logit, since they tend to produce duplicate products in the line, due to the IIA problem. Furthermore preference heterogeneity among consumers is a factor usually neglected, although its representation form has a substantial impact in the design of the product line. We propose a probabilistic choice model that can be used with algorithms that solve the optimal product line design problem, using the share of choices criterion. Our model deals with the IIA problem, by incorporating the similarity among products through the use of a corrective method. In addition, preference heterogeneity among consumers is effectively represented, while the models predictive accuracy is optimized through the use of Stochastic Logarithmic Search and Genetic Algorithms.