María Amparo Vila Miranda

Mining Text Data: Special Features and Patterns

Text mining is an increasingly important research field because of the necessity of obtaining knowledge from the enormous number of text documents available, especially on the Web. Text mining and data mining, both included in the field of information mining, are similar in some sense, and thus it may seem that data mining techniques may be adapted in a straightforward way to mine text. However, data mining deals with structured data, whereas text presents special characteristics and is basically unstructured. In this context, the aims of this paper are three: - To study particular features of text. - To identify the patterns we may look for in text. - To discuss the tools we may use for that purpose.In relation with the third point we overview existing proposals, as well as some new tools we are developing by adapting data mining tools previously developed by our research group.

Applying Genetic Algorithms to the Feature Selection Problem in Information Retrieval

The demand of accuracy and speed in the Information Retrieval processes has revealed the necessity of a good classification of the large collection of documents existing in databases and Web servers. The representation of documents in the vector space model with terms as features offers the possibility of application of Machine Learning techniques. A filter method to select the most relevant features before the classification process is presented in this paper. A Genetic Algorithm (GA) is used as a powerful tool to search solutions in the domain of relevant features. Implementation and some preliminary experiments have been realized. The application of this technique to the vector space model in Information Retrieval is outlined as future work.

Influence of Granularity Level in Fuzzy Functional Dependencies

Relational Databases (R.D) can be considered the most widely used approach to Databases [5]. But in Classical R.D it is not possible to store or treat vague information.

Association Rule Extraction for Text Mining

We present the definition of fuzzy association rules and fuzzy transactions in a text framework. The traditional mining techniques are applied to documents to extract rules. The fuzzy framework allows us to deal with a fuzzy extended Boolean model. Text mining with fuzzy association rules is applied to one of the classical problems in Information Retrieval: query refinement. The extracted rules help users to query the system by showing them a list of candidate terms to refine the query. Different procedures to apply these rules in an automatic and semi-automatic way are also presented.

Using Artificial Neural Network to Define Fuzzy Comparators in FSQL with the Criterion of some Decision-Maker

At present we have a FSQL server available for Oracle© Databases, programmed in PL/SQL. This server allows us to query a Fuzzy or Classical Database with the FSQL language (Fuzzy SQL). The FSQL language is an extension of the SQL language which permits us to write flexible (or fuzzy) conditions in our queries to a fuzzy or traditional database. In this paper we have incorporated a method of ranking fuzzy numbers using Neural Networks to compare fuzzy quantities in FSQL. The main advantage is that any user can to train his own fuzzy comparator for any specific problem We consider that this model satisfies the requirements of Data Mining systems (high-level language, efficiency, certainty, interactivity, etc) and this new level of personal configuration makes the system very useful and flexible.

An Architecture for a Deductive Fuzzy Relational Database

This paper reports on the architecture of a Fuzzy Relational DBMS (FRDBMS) with deduction capabilities, whose main characteristics are: 1) It is built on the basis of a theoretical model for fuzzy relational databases and a theoretical model for logic fuzzy databases; 2) It is implemented entirely on classical RDBMS, using their resources; 3) It conserves all the operations of the host RDBMS and gives them more power, adding new capabilities for dealing with ”fuzzy” and ”intensive” information; 4) It provides a deductive fuzzy language, DFSQL, and a processor which permits the translation of each DFSQL statement into one or more SQL statements, which can be used by the host RDBMS; 5) It offers a relational representaion of the rules that define an intensive table, in such a way that all necessary information to perform deduction is stored in tables. 6) This system needs to interact with a deduction module which performs the computation of intensive tables.

Flexible Management of Essential Construction Tasks Using Fuzzy OLAP Cubes

Abstract The construction sector has significantly evolved in recent decades, in parallel with a huge increase in the amount of data generated and exchanged in any construction project. These data need to be managed in order to complete a successful project in terms of quality, cost and schedule in the the context of a safe project environment while appropriately organising many construction documents. However, the origin of these data is very diverse, mainly due to the sector’s characteristics. Moreover, these data are affected by uncertainty, complexity and diversity due to the imprecise nature of the many factors involved in construction projects. As a result, construction project data are associated with large, irregular and scattered datasets. The objective of this chapter is to introduce an approach based on a fuzzy multi-dimensional model and on line analytical processing (OLAP) operations in order to manage construction data and support the decision-making process based on previous experiences. On one hand, the proposal allows for the integration of data in a common repository which is accessible to users along the whole project’s life cycle. On the other hand, it allows for the establishment of more flexible structures for representing the data of the main tasks in the construction project management domain. The incorporation of this fuzzy framework allows for the management of imprecision in construction data and provides easy and intuitive access to users so that they can make more reliable decisions.

An ontology-based framework for automatic topic detection in multilingual environments

The detection of topics from large textual data volumes is currently a research area, which has many applications in the development of computational systems. A proposed solution for the detection of topics in data mining is the application of clustering methods. This paper presents the application of a new ontology‐based methodology for the automatic topic detection without any previous information based on the use of hierarchical clustering algorithms and a multilingual knowledge base. The approach also includes lexical resources that allow us to enrich the semantics of the analyzed texts. The novelty of this approach consists of the dimensionality reduction of the terms present in the texts by using ontology and the introduction of a method for the creation of a term weight matrix for use in clustering algorithms. With this approach, it is possible to improve automatic topic detection in documents. The proposed methodology was assessed with four datasets (two of them in English and two in Spanish).