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.

/pdf/learning-concept-hierarchies-from-text-corpora-using-formal-2jhbdhcmw9.pdf

Learning concept hierarchies from text corpora using formal concept analysis

Cooperative Information Systems (CoopIS) 2004 International Conference- CoopIS 2004 International Conference (International Conference on Cooperative Information Systems) PC Co-chairs' Message- Keynote- Business Process Optimization- Workflow/Process/Web Services, I- Discovering Workflow Transactional Behavior from Event-Based Log- A Flexible Mediation Process for Large Distributed Information Systems- Exception Handling Through a Workflow- Workflow/Process/Web Services, II- A Flexible and Composite Schema Matching Algorithm- Analysis, Transformation, and Improvements of ebXML Choreographies Based on Workflow Patterns- The Notion of Business Process Revisited- Workflow/Process/Web Services, III- Disjoint and Overlapping Process Changes: Challenges, Solutions, Applications- Untangling Unstructured Cyclic Flows - A Solution Based on Continuations- Making Workflow Models Sound Using Petri Net Controller Synthesis- Database Management/Transaction- Concurrent Undo Operations in Collaborative Environments Using Operational Transformation- Refresco: Improving Query Performance Through Freshness Control in a Database Cluster- Automated Supervision of Data Production - Managing the Creation of Statistical Reports on Periodic Data- Schema Integration/Agents- Deriving Sub-schema Similarities from Semantically Heterogeneous XML Sources- Supporting Similarity Operations Based on Approximate String Matching on the Web- Managing Semantic Compensation in a Multi-agent System- Modelling with Ubiquitous Agents a Web-Based Information System Accessed Through Mobile Devices- Events- A Meta-service for Event Notification- Classification and Analysis of Distributed Event Filtering Algorithms- P2P/Collaboration- A Collaborative Model for Agricultural Supply Chains- FairNet - How to Counter Free Riding in Peer-to-Peer Data Structures- Supporting Collaborative Layouting in Word Processing- A Reliable Content-Based Routing Protocol over Structured Peer-to-Peer Networks- Applications, I- Covering Your Back: Intelligent Virtual Agents in Humanitarian Missions Providing Mutual Support- Dynamic Modelling of Demand Driven Value Networks- An E-marketplace for Auctions and Negotiations in the Constructions Sector- Applications, II- Managing Changes to Engineering Products Through the Co-ordination of Human and Technical Activities- Towards Automatic Deployment in eHome Systems: Description Language and Tool Support- A Prototype of a Context-Based Architecture for Intelligent Home Environments- Trust/Security/Contracts- Trust-Aware Collaborative Filtering for Recommender Systems- Service Graphs for Building Trust- Detecting Violators of Multi-party Contracts- Potpourri- Leadership Maintenance in Group-Based Location Management Scheme- TLS: A Tree-Based DHT Lookup Service for Highly Dynamic Networks- Minimizing the Network Distance in Distributed Web Crawling- Ontologies, DataBases, and Applications of Semantics (ODBASE) 2004 International Conference- ODBASE 2004 International Conference (Ontologies, DataBases, and Applications of Semantics) PC Co-chairs' Message- Keynote- Helping People (and Machines) Understanding Each Other: The Role of Formal Ontology- Knowledge Extraction- Automatic Initiation of an Ontology- Knowledge Extraction from Classification Schemas- Semantic Web in Practice- Generation and Management of a Medical Ontology in a Semantic Web Retrieval System- Semantic Web Based Content Enrichment and Knowledge Reuse in E-science- The Role of Foundational Ontologies in Manufacturing Domain Applications- Intellectual Property Rights Management Using a Semantic Web Information System- Ontologies and IR- Intelligent Retrieval of Digital Resources by Exploiting Their Semantic Context- The Chrysostom Knowledge Base: An Ontology of Historical Interactions- Text Simplification for Information-Seeking Applications- Information Integration- Integration of Integrity Constraints in Federated Schemata Based on Tight Constraining- Modal Query Language for Databases with Partial Orders- Composing Mappings Between Schemas Using a Reference Ontology- Assisting Ontology Integration with Existing Thesauri

On the move to meaningful Internet systems 2004 : CoopIS, DOA, and ODBASE : OTM confederated international conferences, CoopIS, DOA, and ODBASE 2004, Agia Napa, Cyprus, October 25-29, 2004 : proceedings (part II)

In recent years several measures for the gold standard based evaluation of ontology learning were proposed. They can be distinguished by the layers of an ontology (e.g. lexical term layer and concept hierarchy) they evaluate. Judging those measures with a list of criteria we show that there exist some measures sufficient for evaluating the lexical term layer. However, existing measures for the evaluation of concept hierarchies fail to meet basic criteria. This paper presents a new taxonomic measure which overcomes the problems of current approaches.

/pdf/on-how-to-perform-a-gold-standard-based-evaluation-of-3ngulq5obp.pdf

On how to perform a gold standard based evaluation of ontology learning

It is believed that transactions belong to key services of componentbased middleware. However, surprisingly, there is no general agreement on how the component-based middleware support for transactions should look like. In this paper we exploit our experiences with Jironde [4], a flexible framework that extends the Fractal component model [1, 2] with transactions via a set of transactional controllers that manage transactions on behalf of a component. Then we identify several key architectural and technical issues related to enhancing component-based middleware with transactions. We investigate how current technologies address theses issues, and the suitability of current standards to support transaction processing in component-based middleware. 1 Transactions and Components: Architectural Issues Different component models deal with component’s participation in a transaction differently. In the explicit transaction participation, the scenario of involving a component C to a transaction t essentially consists of the three steps as follows: 1) C is registered to t. 2) A client invokes various operations on C. 3) At the time of t’s validation, the transaction manager invokes specific methods of the registered C’s interfaces. These (callback) methods must be implemented by the transactional components. With the implicit transaction participation, components are not obliged to implement any functionality related to transactions. Any time C is visited by a transaction t, the transaction manager of the container keeps all necessary information to manage atomicity, concurrency control, and recovery. Different component standards deal with component participation in transactions differently. CCM use the explicit transaction participation, COM+ uses the implicit one, while EJB mix both. Component-unaware transactions manipulate data without any knowledge on whether they are organized or related to components. For component-aware transactions, components are the data they manipulate with. We believe that all the CCM, EJB, and COM+ transactions are component-unaware. A component is transaction-unaware if its code does not use any transactional primitives and is not therefore in any way dependent on any transactional standard, while a transactionaware component is the opposite. Component’s awareness of transactions reflects the implicit/explicit transaction participation but includes also some hidden expectations of the component design. Transaction-unaware container does not deal with On Enhancing Component-Based Middleware with Transactions 2 transactions, which are managed at the application level instead. Transaction-aware container provides some transaction management features, such as containerdemarcation, transaction propagation, concurrency control, etc. EJB, CCM and COM+ are examples that provide such containers. 2 Transactions and Components: Technical Issues The technical issues related to transactional component include concurrency control, recovery, and transaction context propagation. As for concurrency control, the current technologies either use a simple read/write model or do not allow any concurrent access to component instances at all. It would be beneficial to exploit the concurrency potential of components by e.g. the use of conflict matrixes defined on all the methods of all implemented interfaces. Both concurrency control and recovery of components reflect the architectural patterns presented in the previous section. Another important issue is the transaction context propagation. There are several options of how to specify a transaction propagation policy, as well as whether to define it either during the component’s development or during its deployment. A last issue is to support the definition of new propagation policies, like JOTDF [5] does.

On The Move to meaningful Internet systems 2003 : OTM 2003 Workshops : OTM Confederated International Workshops : HCI-SWWA, IPW, JTRES, WORM, WMS, and WRSM 2003, Catania, Sicily, Italy, November 3-7, 2003 : proceedings

Domain ontologies very rarely model verbs as relations holding between concepts. However, the role of the verb as a central connecting element between concepts is undeniable. Verbs specify the interaction between the participants of some action or event by expressing relations between them. In parallel, it can be argued from an ontology engineering point of view that verbs express a relation between two classes that specify domain and range. The work described here is concerned with relation extraction for ontology extension along these lines. We describe a system (RelExt) that is capable of automatically identifying highly relevant triples (pairs of concepts connected by a relation) over concepts from an existing ontology. RelExt works by extracting relevant verbs and their grammatical arguments (i.e. terms) from a domain-specific text collection and computing corresponding relations through a combination of linguistic and statistical processing. The paper includes a detailed description of the system architecture and evaluation results on a constructed benchmark. RelExt has been developed in the context of the SmartWeb project, which aims at providing intelligent information services via mobile broadband devices on the FIFA World Cup that will be hosted in Germany in 2006. Such services include location based navigational information as well as question answering in the football domain.

/pdf/relext-a-tool-for-relation-extraction-from-text-in-ontology-11t0pc9jlc.pdf

RelExt : a tool for relation extraction from text in ontology extension

Ontologies in current computer science parlance are computer based resources that represent shared conceptualizations for a specific domain. This paper first introduces ontologies in general and subsequently, in particular, shortly outlines the DOGMA ontology leaning approach. The paper also introduces the reader in the field of Knowledge Discovery in Text before, in the main part, work in progress is described and experimentally evaluated. It concerns a potential method to automatically extract concepts and conceptual relationships from texts. Preliminary outcomes are presented based on the clustering of nominal terms and prepositional phrases according to co-occurrence frequencies in the verb-object syntactic context.

Discovering Knowledge in Texts for the learning of DOGMA-inspired ontologies

We report on an a set of experiments carried out in the context of the Flemish OntoBasis project. Our purpose is to extract semantic relations from text corpora in an unsupervised way and use the output as preprocessed material for the construction of ontologies from scratch. The experiments are evaluated in a quantitative and ”impressionistic” manner.

Automatic Initiation of an Ontology

Unsupervised Text Mining for the learning of DOGMA-inspired Ontologies

Our purpose is to present a method to lexically evaluate the results of extracting in an unsupervised way material from text corpora to build ontologies. We have worked on a legal corpus (EU VAT directive) consisting of 43K words. The unsupervised text miner has produced a set of triples. These are to be used as preprocessed material for the construction of ontologies from scratch. A quantitative scoring method (coverage, accuracy, recall and precision metrics resulting in a 38.68%, 52.1%, 9.84% and 75.81% scores respectively) has been defined and applied.

/pdf/lexically-evaluating-ontology-triples-generated-2892nzomqp.pdf

Lexically evaluating ontology triples generated automatically from texts

Ontologies in current computer science parlance are computer based resources that represent agreed domain semantics. This paper first introduces ontologies in general and subsequently, in particular, shortly outlines the DOGMA ontology engineering approach that separates “atomic” conceptual relations from “predicative” domain rules. In the main part of the paper, we describe and experimentally evaluate work in progress on a potential method to automatically derive the atomic conceptual relations mentioned above from a corpus of English medical texts. Preliminary outcomes are presented based on the clustering of nouns and compound nouns according to co-occurrence frequencies in the subject-verb-object syntactic context.

Marie-Laure Reinberger

Papers

Discovering Knowledge in Texts for the learning of DOGMA-inspired ontologies

Automatic Initiation of an Ontology

Unsupervised Text Mining for the learning of DOGMA-inspired Ontologies

Lexically evaluating ontology triples generated automatically from texts

Mining for Lexons: Applying Unsupervised Learning Methods to Create Ontology Bases