Catálogo de publicaciones - libros

Frequent Pattern Mining (FPM) is a very powerful paradigm for mining informative and useful patterns in massive, complex datasets. In this paper we propose the Data Mining Template Library, a collection of generic containers and algorithms for FPM, as well as persistency and database management classes. DMTL provides a systematic solution to a whole class of common FPM tasks like itemset, sequence, tree and graph mining. DMTL is extensible, scalable, and high-performance for rapid response on massive datasets. Our experiments show that DMTL is competitive with special purpose algorithms designed for a particular pattern type, especially as database sizes increase.

The “Semantic Mirrors Method” (Dyvik, 1998) is a means for automatic derivation of thesaurus entries from a word-aligned parallel corpus. The method is based on the construction of lattices of linguistic features. This paper models the Semantic Mirrors Method with Formal Concept Analysis. It is argued that the method becomes simpler to understand with the help of FCA. This paper then investigates to what extent the Semantic Mirrors Method is applicable if the linguistic resource is not a high quality parallel corpus but, instead, a medium quality bilingual dictionary. This is a relevant question because medium quality bilingual dictionaries are freely available whereas high quality parallel corpora are expensive and difficult to obtain. The analysis shows that by themselves, bilingual dictionaries are not as suitable for the Semantic Mirrors Method but that this can be improved by applying conceptual exploration. The combined method of conceptual exploration and Semantic Mirrors provides a useful toolkit specifically for smaller size bilingual resources, such as ontologies and classification systems. The last section of this paper suggests that such applications are of interest in the area of ontology engineering.

An interesting problem associated with the World Wide Web (Web) is the definition and delineation of so called Web communities. The Web can be characterized as a directed graph whose nodes represent Web pages and whose edges represent hyperlinks. An authority is a page that is linked to by high quality hubs, while a hub is a page that links to high quality authorities. A Web community is a highly interconnected aggregate of hubs and authorities. We define a community core to be a maximally connected bipartite subgraph of the Web graph.

We observe that the web subgraph can be viewed as a formal context and that web communities can be modeled by formal concepts. Additionally, the notions of hub and authority are captured by the extent and intent, respectively, of a concept. Though Formal Concept Analysis (FCA) has previously been applied to the Web, none of the FCA based approaches that we are aware of consider the link structure of the Web pages. We utilize notions from FCA to explore the community structure of the Web graph. We discuss the problem of utilizing this structure to locate and organize communities in the form of a knowledge base built from the resulting concept lattice and discuss methods to reduce the complexity of the knowledge base by coalescing similar Web communities. We present preliminary experimental results obtained from real Web data that demonstrate the usefulness of FCA for improving Web search.

Automatic attribute selection is a critical step when using Formal Concept Analysis (FCA) in a free text document retrieval framework. Optimal attributes as document descriptors should produce smaller, clearer and more browsable concept lattices with better clustering features. In this paper we focus on the automatic selection of noun phrases as document descriptors to build an FCA-based IR framework. We present three different phrase selection strategies which are evaluated using the and the evaluation measures. Noun phrases are shown to produce lattices with good clustering properties, with the advantage (over simple terms) of being better intensional descriptors from the user’s point of view.

In this paper we report on practical information visualization aspects of Conceptual Knowledge Processing (CKP), realizing and illustrating Wille’s “conceptual landscapes” in the context of developing a conceptual information system to determine surfing conditions on the South Coast of New South Wales, Australia. This novel application illustrates some (if not all) of Wille’s CKP tasks: exploring, searching, recognizing, identifying, analyzing, investigating, deciding, restructuring and memorizing (all but improving). It does this by concentrating on combining an information landscape with maps of the physical world.

This paper reports on ongoing work to use Formal Concept Analysis as an auxiliary tool in understanding and visualising the wealth of data produced by lexical-resource building as embodied in the construction of FrameNet, a database to capture the syntax and semantics of language use in Frame Linguistics. We present proof of the abundance of concept lattices both in the theory of frames and in its present day incarnation, the FrameNet resource, with contributions that range from data-visualisation to the fine-tuning of some lexico-theoretical concepts better understood in terms of Formal Concept Analysis.

A key difficulty in the maintenance and evolution of complex software systems is to recognize and understand the implicit dependencies that define contracts that must be respected by changes to the software. Formal Concept Analysis is a well-established technique for identifying groups of elements with common sets of properties. We have successfully applied FCA to complex software systems in order to automatically discover a variety of different kinds of implicit, recurring sets of dependencies amongst design artifacts. In this paper we describe our approach, outline three case studies, and draw various lessons from our experiences. In particular, we discuss how our approach is applied iteratively in order to draw the maximum benefit offered by FCA.

Information technology of today is often concerned with information that is not only large in quantity but also complex in structure. Understanding this structure is important in many domains – many quantitative approaches such as data mining have been proposed to address this issue. This paper presents a conceptual approach based on Formal Concept Analysis. Using software source code as an example of a complex structure we present a framework for conceptually analysing relational structures. In our framework, a browsable space of sub-contexts is automatically derived from a database of relations augmented by a rule engine and schema information. Operations are provided for the user to navigate between sub-contexts. We demonstrate how the use of these operations can lead to quick identification of an area of software source code that establishes an unecessary dependency between software parts.

This paper extends standard help system technology to demonstrate the suitability of Formal Concept Analysis in displaying, searching and navigating help content. The paper introduces a method for building suitable scales directly from the help system index by computing a keyword extension set. The keyword extension technique is generalisable in any document collection where a hand-crafted index of terms is available.

This paper illustrates how a new clustering process dedicated to the analysis of anecdotal reports of aviation incidents has been designed and tested thanks to an FCA tool called . Special attention has been given to the necessary transcription of the data from the initial relational database to an FCA context. The graphical interface for , which has been specially implemented for this study, is also presented.

The study presented in this paper validates the process adopted and highlights the use of FCA to help the expert to mine the database without previous knowledge of the searched concepts. The work brought original ideas to the aviation safety community by the development of an incident model and the notion of scenario. For the FCA community, one interesting aspect of this work lies on the use of a first-order context (given by a relational database) and its translation into a classical context.