Atanas Kiryakov

Semantic annotation, indexing, and retrieval

The Semantic Web realization depends on the availability of a critical mass of metadata for the web content, associated with the respective formal knowledge about the world. We claim that the Semantic Web, at its current stage of development, is in a state of a critical need of metadata generation and usage schemata that are specific, well-defined and easy to understand. This paper introduces our vision for a holistic architecture for semantic annotation, indexing, and retrieval of documents with regard to extensive semantic repositories. A system (called KIM), implementing this concept, is presented in brief and it is used for the purposes of evaluation and demonstration. A particular schema for semantic annotation with respect to real-world entities is proposed. The underlying philosophy is that a practical semantic annotation is impossible without some particular knowledge modelling commitments. Our understanding is that a system for such semantic annotation should be based upon a simple model of real-world entity classes, complemented with extensive instance knowledge. To ensure the efficiency, ease of sharing, and reusability of the metadata, we introduce an upper-level ontology (of about 250 classes and 100 properties), which starts with some basic philosophical distinctions and then goes down to the most common entity types (people, companies, cities, etc.). Thus it encodes many of the domain-independent commonsense concepts and allows straightforward domain-specific extensions. On the basis of the ontology, a large-scale knowledge base of entity descriptions is bootstrapped, and further extended and maintained. Currently, the knowledge bases usually scales between 10^5 and 10^6 descriptions. Finally, this paper presents a semantically enhanced information extraction system, which provides automatic semantic annotation with references to classes in the ontology and to instances. The system has been running over a continuously growing document collection (currently about 0.5 million news articles), so it has been under constant testing and evaluation for some time now. On the basis of these semantic annotations, we perform semantic based indexing and retrieval where users can mix traditional information retrieval (IR) queries and ontology-based ones. We argue that such large-scale, fully automatic methods are essential for the transformation of the current largely textual web into a Semantic Web.

OWLIM – a pragmatic semantic repository for OWL

OWLIM is a high-performance Storage and Inference Layer (SAIL) for Sesame, which performs OWL DLP reasoning, based on forward-chaining of entilement rules. The reasoning and query evaluation are performed in-memory, while in the same time OWLIM provides a reliable persistence, based on N-Triples files. This paper presents OWLIM, together with an evaluation of its scalability over synthetic, but realistic, dataset encoded with respect to PROTON ontology. The experiment demonstrates that OWLIM can scale to millions of statements even on commodity desktop hardware. On an almost-entry-level server, OWLIM can manage a knowledge base of 10 million explicit statements, which are extended to about 19 millions after forward chaining. The upload and storage speed is about 3,000 statement/sec. at the maximal size of the repository, but it starts at more than 18,000 (for a small repository) and slows down smoothly. As it can be expected for such an inference strategy, delete operations are expensive, taking as much as few minutes. In the same time, a variety of queries can be evaluated within milliseconds. The experiment shows that such reasoners can be efficient for very big knowledge bases, in scenarios when delete operations should not be handled in real-time.

KIM – a semantic platform for information extraction and retrieval

The KIM platform provides a novel Knowledge and Information Management framework and services for automatic semantic annotation, indexing, and retrieval of documents. It provides a mature and semantically enabled infrastructure for scalable and customizable information extraction (IE) as Our understanding is that a system for semantic annotation should be based upon a simple model of real-world entity concepts, complemented with quasi-exhaustive instance knowledge. To ensure efficiency, easy sharing, and reusability of the metadata we introduce an upper-level ontology. Based on the ontology, a large-scale instance base of entity descriptions is maintained. The knowledge resources involved are handled by use of state-of-the-art Semantic Web technology and standards, including RDF(S) repositories, ontology middleware and reasoning. From a technical point of view, the platform allows KIM-based applications to use it for automatic semantic annotation, for content retrieval based on semantic queries, and for semantic repository access. As a framework, KIM also allows various IE modules, semantic repositories and information retrieval engines to be plugged into it. This paper presents the KIM platform, with an emphasis on its architecture, interfaces, front-ends, and other technical issues.

KIM: semantic annotation platform

The KIM platform provides a novel Knowledge and Information Management infrastructure and services for automatic semantic annotation, indexing, and retrieval of documents. It provides mature infrastructure for scaleable and customizable information extraction (IE) as well as annotation and document management, based on GATE. In order to provide basic level of performance and allow easy bootstrapping of applications, KIM is equipped with an upper-level ontology and a knowledge base providing extensive coverage of entities of general importance. The ontologies and knowledge bases involved are handled using cutting edge Semantic Web technology and standards, including RDF(S) repositories, ontology middleware and reasoning. From technical point of view, the platform allows KIM-based applications to use it for automatic semantic annotation, content retrieval based on semantic restrictions, and querying and modifying the underlying ontologies and knowledge bases. This paper presents the KIM platform, with emphasize on its architecture, interfaces, tools, and other technical issues.

Ontology Versioning and Change Detection on the Web

To effectively use ontologies on the Web, it is essential that changes in ontologies are managed well. This paper analyzes the topic of ontology versioning in the context of the Web by looking at the characteristics of the version relation between ontologies and at the identification of online ontologies. Then, it describes the design of a web-based system that helps users to manage changes in ontologies. The system helps to keep different versions of web-based ontologies interoperable, by maintaining not only the transformations between ontologies, but also the conceptual relation between concepts in different versions. The system allows ontology engineers to compare versions of ontology and to specify these conceptual relations. For the visualization of differences, it uses an adaptable rule-based mechanism that finds and classifies changes in RDF-based ontologies.

OWLIM: A family of scalable semantic repositories

An explosion in the use of RDF for representing information about resources has driven the requirements for Web-scale server systems that can store and process huge quantities of data, and furthermore provide powerful data access and mining functionalities. This paper describes OWLIM, a family of semantic repositories that provide storage, inference and novel data-access features delivered in a scalable, resilient, industrial-strength platform.

Experiments with geographic knowledge for information extraction

Here we present work on using spatial knowledge in conjunction with information extraction (IE). Considerable volume of location data was imported in a knowledge base (KB) with entities of general importance used for semantic annotation, indexing, and retrieval of text. The Semantic Web knowledge representation standards are used, namely RDF(S). An extensive upper-level ontology with more than two hundred classes is designed. With respect to the locations, the goal was to include the most important categories considering public and tasks not specially related to geography or related areas. The locations data is derived from number of publicly available resources and combined to assure best performance for domain-independent named-entity recognition in text. An evaluation and comparison to high performance IE application is given.

Tracking Changes in RDF(S) Repositories

The real-world knowledge management applications require features such as versioning and fine-grained access control. Each of them raises the issue of tracking the changes in a knowledge base. Important part of the research presented is the definition of a formal model for tracking changes in graph-based data models. It was used in the ontology middleware module developed under the On-To-Knowledge project as an extension of the Sesame RDF(S) repository. This paper is further development of the results reported in [5].

OntoMap: portal for upper-level ontologies

Currently the evaluation of the feasibility of general-purposeontologies and upper-level models is expensive mostly because oftechnical problems such as different representation formalisms andterminologies used. Additionally, there are no formal mappingsbetween the upper-level ontologies that could ease any kind ofstudies and comparisons. We present the OntoMap Project(http://www.OntoMap.org), a project with the pragmatic goal tofacilitate the access, understanding, and reuse of such resources.A semantic framework on the conceptual level is implemented that issmall and easy enough to be learned on-the-fly. We tried to designthe framework so that it captures most of the semantics usuallyencoded in upper-level models. Technically, OntoMap is a web-siteproviding access to several upper-level ontologies and manualmapping between them.

OPJK into PROTON: legal domain ontology integration into an upper-level ontology

The SEKT Project aims at developing and exploiting the knowledge technologies which underlie the Next Generation Knowledge Management, connecting complementary know-how of key European centers in three areas: Ontology Management Technology, Knowledge Discovery and Human Language Technology. This paper describes the development of PROTON, an upper-level ontology developed by Ontotext, and of the Ontology of Professional Judicial Knowledge (OPJK), modeled by a team of legal experts form the Institute of Law and Technology (IDT-UAB) for the Iuriservice prototype (a webbased intelligent FAQ for the Spanish judges on their first appointment designed by iSOCO). The paper focuses on the work done towards the integration of the OPJK built using a middle-out strategy into the system and top modules of PROTON, illustrating the flexibility of this independent upper-level ontology.