scholarly journals Learning Concept Hierarchies from Text Corpora using Formal Concept Analysis

2005 ◽  
Vol 24 ◽  
pp. 305-339 ◽  
Author(s):  
P. Cimiano ◽  
A. Hotho ◽  
S. Staab

We present a novel approach to the automatic acquisition of taxonomies or concept hierarchies from a text corpus. The approach is based on Formal Concept Analysis (FCA), a method mainly used for the analysis of data, i.e. for investigating and processing explicitly given information. We follow Harris' distributional hypothesis and model the context of a certain term as a vector representing syntactic dependencies which are automatically acquired from the text corpus with a linguistic parser. On the basis of this context information, FCA produces a lattice that we convert into a special kind of partial order constituting a concept hierarchy. The approach is evaluated by comparing the resulting concept hierarchies with hand-crafted taxonomies for two domains: tourism and finance. We also directly compare our approach with hierarchical agglomerative clustering as well as with Bi-Section-KMeans as an instance of a divisive clustering algorithm. Furthermore, we investigate the impact of using different measures weighting the contribution of each attribute as well as of applying a particular smoothing technique to cope with data sparseness.

2015 ◽  
Vol 24 (02) ◽  
pp. 1540006
Author(s):  
Michel Plantié ◽  
Michel Crampes

Concept Hierarchies and Formal Concept Analysis (FCA) are theoretically well grounded. They rely on line diagrams called Galois lattices for visualizing and analysing object-attribute sets. Galois lattices are visually seducing and conceptually rich for experts. However they present important drawbacks due to their concept oriented overall structure: analysing what they show is difficult for non experts, navigation is cumbersome, interaction is poor, and scalability is a deep bottleneck for visual interpretation even for experts. In this paper we introduce semantic probes as a means to overcome many of these problems and extend usability and application possibilities of traditional FCA visualization methods. Semantic probes are visual user centred objects which extract and organize reduced Galois sub-hierarchies. They are simpler, clearer, and they provide a better navigation support through a rich set of interaction possibilities. Since probe driven sub-hierarchies are limited to users' focus, scalability is under control and interpretation is facilitated. After some successful experiments, several applications are being developed with the remaining problem of finding a compromise between simplicity and conceptual expressivity.


Author(s):  
Adriana M. Guimaraes de Farias ◽  
Marcos E. Cintra ◽  
Angelica C. Felix ◽  
Danniel L. Cavalcante

Public security has always been an important research topic. In this sense, machine learning algorithms have been used to extract knowledge from criminal databases, which usually maintain records in order to generate statistics. The automatic extraction of knowledge from such databases allows the improvement and planning of strategies to prevent and combat crimes. Accordingly, in this work different models related to public security are presented. Such models are based on clustering algorithms, on the analysis of formal concept techniques, and on the analysis of crime record data collected in the city of Mossoro, Brazil. The two types of models generated are: (i) concept lattices with crime patterns; (ii) criminal hot spot maps. We also produced a ranking of dangerousness for neighbourhoods of Mossoro. The Fuzzy K-Means clustering algorithm was used to obtain criminal hot spots, which indicate locations with high crime incidence. Formal concept analysis was used for extracting visual models describing patterns that characterize criminal activities. Such models have the form of conceptual lattices that provide graphical displays which can be used for defining strategies to combat and prevent crime. The models were first empirically evaluated and then analysed by public security experts, who provided positive feedback for their practical use. The advantages of the automatically generated models presented in this paper are many, including the short time to produce such models, the variety of different models that can be generated for specific regions and periods of days, months, or years, the graphical characteristic of such models that allow a fast analysis of them, as well as the use of large amounts of data, which are infeasible activities to be done by human experts.


Author(s):  
Franc Grootjen ◽  
Theo van der Weide

To effectively use and exchange information among AI systems, a formal specification of the representation of their shared domain of discourse—called an ontology—is indispensable. In this chapter we introduce a special kind of knowledge representation based on a dual view on the universe of discourse and show how it can be used in human activities such as searching, in-depth exploration and browsing. After a formal definition of dualistic ontologies we exemplify this definition with three different (well known) kinds of ontologies, based on the vector model, on formal concept analysis and on fuzzy logic respectively. The vector model leads to concepts derived by latent semantic indexing using the singular value decomposition. Both the set model and the fuzzy-set model lead to formal concept analysis, in which the fuzzy-set model is equipped with a parameter that controls the fine-graining of the resulting concepts. We discuss the relation between the resulting systems of concepts. Finally, we demonstrate the use of this theory by introducing the dual search engine. We show how this search engine can be employed to support the human activities addressed above.


2021 ◽  
Vol 179 (3) ◽  
pp. 295-319
Author(s):  
Longchun Wang ◽  
Lankun Guo ◽  
Qingguo Li

Formal Concept Analysis (FCA) has been proven to be an effective method of restructuring complete lattices and various algebraic domains. In this paper, the notion of contractive mappings over formal contexts is proposed, which can be viewed as a generalization of interior operators on sets into the framework of FCA. Then, by considering subset-selections consistent with contractive mappings, the notions of attribute continuous formal contexts and continuous concepts are introduced. It is shown that the set of continuous concepts of an attribute continuous formal context forms a continuous domain, and every continuous domain can be restructured in this way. Moreover, the notion of F-morphisms is identified to produce a category equivalent to that of continuous domains with Scott continuous functions. The paper also investigates the representations of various subclasses of continuous domains including algebraic domains and stably continuous semilattices.


2013 ◽  
Vol 760-762 ◽  
pp. 1708-1712
Author(s):  
Ying Fang Li ◽  
Ying Jiang Li ◽  
Yan Li ◽  
Yang Bo

At present, as the number of web services resources on the network drastically increased, how to quickly and efficiently find the needed services from publishing services has become a problem to resolve. Aiming at the problems of low efficiency in service discovery of traditional web service, the formal concept analysis ( FCA) is introduced into the semantic Web service matching, and a Matching Algorithm based semantic web service is proposed. With considering the concept of limited inheritance,this method introduces the concept of limited inheritance to the semantic similarity calculation based on the concept lattice. It is significant in enhancing the service function matching in practical applications through adjust the calculation.


Sign in / Sign up

Export Citation Format

Share Document