Javier Segovia

Web usage mining project for improving web-based learning sites

Despite the great success of data mining being applied for personalization in web environments, it has not yet been massively applied in the e-learning domains. In this paper, we outline a web usage mining project which has been initiated in University of Cantabria. The aim of this project is to develop tools which let us improve its Web-based learning environment in two main aspects: the first that the teacher obtains information which allows him to evaluate the learning process and the second that the student feels supported in this task.

Intelligent exploration of the web

Creation and Representation of Web Resources.- Structure Analysis and Generation for Internet Documents.- A Fuzzy System for the Web Page Representation.- Flexible Representation and Retrieval of Web Documents.- Information Retrieval.- Intelligent Information Retrieval on the Web.- Internet as a Challenge to Fuzzy Querying.- Internet Search Based on Text Intuitionistic Fuzzy Similarity.- Content-Based Fuzzy Search in a Multimedia Web Database.- Self-Organizing Maps for Interactive Search in Document Databases.- Methods for Exploratory Cluster Analysis.- Textual Information Retrieval with User Profiles Using Fuzzy Clustering and Inferencing.- Intelligent Clustering as Source of Knowledge for Web Dialogue Manager in an Information Retrieval System.- Document Clustering Using Tolerance Rough Set Model Its Application to Information Retrieval.- Improving Web Search by the Identification of Contextual Information.- Intelligent Internet-Based Multiagent Systems.- Neural Agent for Text Database Discovery.- Intelligent Web Agents that Learn to Retrieve and Extract Information.- Accurately and Reliably Extracting Data from the Web: A Machine Learning Approach.- Web Browsing Using Machine Learning on Text Data.- Retrieval of Semistructured Web Data.- Intelligent Retrieval of Hypermedia Documents.- Bootstrapping an Ontology-Based Information Extraction System.- Web Data Mining and Use.- Intelligent Web Mining.- A Neural Net Approach to Data Mining: Classification of Users to Aid Information Management.- Web-Based Expert Systems: Information Clients versus Knowledge Servers.

Toward data mining engineering: A software engineering approach

The number, variety and complexity of projects involving data mining or knowledge discovery in databases activities have increased just lately at such a pace that aspects related to their development process need to be standardized for results to be integrated, reused and interchanged in the future. Data mining projects are quickly becoming engineering projects, and current standard processes, like CRISP-DM, need to be revisited to incorporate this engineering viewpoint. This is the central motivation of this paper that makes the point that experience gained about the software development process over almost 40 years could be reused and integrated to improve data mining processes. Consequently, this paper proposes to reuse ideas and concepts underlying the IEEE Std 1074 and ISO 12207 software engineering model processes to redefine and add to the CRISP-DM process and make it a data mining engineering standard.

E-Commerce and Intelligent Methods

1. Foundations of Electronic Commerce.- Barriers to Global Electronic Commerce.- Foundations of Electronic Data Interchange.- Some Legal Aspects of Electronic Commerce.- 2. Neural Networks.- Competitive Neural Networks for Customer Choice Models.- CRM in e-Business: a Clients Life Cycle Model Based on a Neural Network.- Customer Relationship Management Systems: The Application of Fuzzy ART Neural Network.- Characterizing and Segmenting the Online Customer Market Using Neural Netwoks.- Data Mining for Diverse E-Commerce Applications.- Extreme Sample Classification and Credit Card Fraud Detection.- 3. Evolutionary Programming.- A Review of Evolutionary Algorithms for E-Commerce.- Artificial Adaptive Market Traders Based in Genetic Algorithms for a Stock Market Simulator.- Data Mining in Marketing Using Bayesian Networks and Evolutionary Programming.- Improving User Profiles for E-Commerce by Genetic Algorithms.- 4. Fuzzy Logic.- Automatic Web User Profiling and Personalization Using Robust Fuzzy Relational Clustering.- Fuzzy Quantifiable Trust in Secure E-Commerce.- Fuzzy Similarity in E-Commerce Domains.- 5. CBR and Agents.- Agencies of Agents for Logistic Applications.- Intelligent Customer Support for Product Selection with Case-Based Reasoning.- Mobile Agent Based Auctionlike Negotiation in Internet Retail Commerce.- Author Index.

Beyond formal processes: augmenting workflow with group interaction techniques

The main scope of workflow systems has been the automation of <italic>formal</italic> procedures in the workplace. On the other hand, Communication and Group Support systems have addressed the <italic>informal</italic> aspects of organizational interactions. We argue that the <italic>formal</italic> versus <italic>informal</italic> separation is artificial and a cause of systems ineffectiveness. This paper proposes an approach to increase mutual awareness when integrating support for workflow systems and group interaction techniques.

Towards a Methodology for Data Mining Project Development: The Importance of Abstraction

Standards such as CRISP-DM, SEMMA, PMML, are making data mining processes easier. Nevertheless, up to date, projects are being developed more as an art than as a science making it difficult to understand, evaluate and compare results as there is no standard methodology. In this chapter, we make a proposal for such a methodology based on RUP and CRISP-DM and concentrate on the project conception phase for determining a feasible project plan.

Interactive Evolutionary Computation algorithms applied to solve Rastrigin test functions

this paper presents a new approach to interactive evolutionary computation that helps the user in the difficult task of finding an optimal solution between multiple possibilities. There are several ways of applying algorithms in interactive evolutionary computation; in this paper we explain three of them in order to make an experimental comparative study. Proceeding with a main goal of solving complex problems as fast as possible, we take the Rastrigin test function as a benchmark and it is executed with the three algorithms described. The aim is to show clearly the results of the algorithms in terms of solution quality and number of iterations. The results clearly show that the use of the proposed method based on chromosome learning heuristics works well even for non Interactive Evolutionary Computation frameworks.

RTCS: a Reactive with Tags Classifier System

In this work, a new Classifier System is proposed (CS). The system, a Reactive with Tags Classifier System (RTCS), is able to take into account environmental situations in intermediate decisions. CSs are special production systems, where conditions and actions are codified in order to learn new rules by means of Genetic Algorithms (GA). The RTCS has been designed to generate sequences of actions like the traditional classifier systems, but RTCS also has the capability of chaining rules among different time instants and reacting to new environmental situations, considering the last environmental situation to take a decision. In addition to the capability to react and generate sequences of actions, the design of a new rule codification allows the evolution of groups of specialized rules. This new codification is based on the inclusion of several bits, named tags, in conditions and actions, which evolve by means of GA. RTCS has been tested in robotic navigation. Results show the suitability of this approximation to the navigation problem and the coherence of tag values in rules classification.

Genetic algorithms for the generation of models with micropopulations

The present article puts forward a method for an interactive model generation through the use of Genetic Algorithms applied to small populations. Micropopulations actually worsen the problem of the premature convergence of the algorithm, since genetic diversity is very limited. In addition, some key factors, which modify the changing likelihood of alleles, cause the likelihood of premature convergence to decrease. The present technique has been applied to the design of 3D models, starting from generic and standard pieces, using objective searches and searches with no defined objective.

Extending UML for Modeling Data Mining Projects (DM-UML)

Existing Data Mining process models propose one way or another of developing projects in a structured manner, trying to reduce their complexity through effective project management. It is well-known in any engineering environment that one of the management tasks that helps to reduce project problems is systematic project documentation, but few of the existing Data Mining processes propose their documentation. Furthermore, these few remark the need of producing documentation at each phase as an input for the next, but they don’t show how to do it. On the other hand, in the literature there are examples of UML extensions for data mining projects, but they always focus on the model implementation side and fail to take into account the remainder of the process. In this paper, we present an extension of the UML modeling language for data mining projects (DM-UML) covering all the documentation needs for a project conforming to a standard process, namely CRISP-DM, ranging from business understanding to deployment. We also show an example of a real application of the proposed DM-UML modeling. The result of this approach is that, besides the advantages of having an standardized way of producing the documentation, it clearly constitutes a very useful and transparent tool for modeling and connecting the business understanding or modeling phase with the remainder of the project right through to deployment, as well as a way of facilitating the communication with the nontechnical stakeholders involved in the project, problems which have always been an open question in data mining.