Geir Solskinnsbakk

A hybrid approach to constructing tag hierarchies

Folksonomies are becoming increasingly popular. They contain large amounts of data which can be mined and utilized for many tasks like visualization, browsing, information retrieval etc. An inherent problem of folksonomies is the lack of structure. In this paper we present an unsupervised approach for generating such structure based on a combination of association rule mining and the underlying tagged material. Using the underlying tagged material we generate a semantic representation of each tag. The semantic representation of the tags is an integral component of the structure generated. The experiment presented in this paper shows promising results with tag structures that correspond well with human judgment.

Mining tag similarity in folksonomies

Folksonomies are becoming increasingly popular, both among users who find them simple and intuitive to use, and scientists as interesting research objects. Folksonomies can be viewed as large informal sources of semantics. Harnessing the semantics for search or concept extraction requires us to be able to recognize linguistic similarity between tags. In this paper we propose an approach that uses a combination of morpho-syntactic and semantic similarity measures without using any external linguistic resources to mine tag pairs that can be reduced to base tags. Our approach is based on the Levenshtein distance for morpho-syntactic similarity and tag signatures for semantic similarity. The evaluation of our approach, based on a data set crawled from Delicious, shows that we are able to recognize a wide range of linguistic variations with high quality.

Editorial: Quality of hierarchies in ontologies and folksonomies

Ontologies have been a hot research topic for the recent decade and have been used for many applications such as information integration, semantic search, knowledge management, etc. Manual engineering of ontologies is a costly process and automatic ontology engineering lacks in precision. Folksonomies have recently emerged as another hot research topic and several research efforts have been made to extract (lightweight) ontologies automatically from folksonomy data. Due to the high cost of manual ontology engineering and the lack of precision in automatic ontology engineering it is important that we are able to evaluate the structure of the ontology. Detection of problems with the suggested ontology at an early stage can, especially for manually engineered ontologies, be cost saving. In this paper we present an approach to evaluate the quality of hierarchical relations in ontologies and folksonomy based structures. The approach is based on constructing shallow semantic representations of the ontology concepts and folksonomy tags. We specify four hypotheses regarding the semantic representations and different quality aspects of the hierarchical relations and perform an evaluation on two different data sets. The results of the evaluation confirm our hypotheses.

Ontological Profiles as Semantic Domain Representations

Ontologies are conceptualizations of some domain, defining its concepts and their relationships. An interesting question is how do we relate the ontological concepts to the general vocabulary of the domain?In this paper we present the concept of ontological profiles, what they are, how they are constructed, and how they may be used. We propose that the ontological profile is a link between the vocabulary of the domain and its conceptual specification given by the ontology. This means that the ontological profile can be tailored to a specific document collection, reflecting the vocabulary actually used. Finally we demonstrate how the ontological profile may be utilized for ontology-driven search.

Concept Signatures and Semantic Drift

Ontology evolution is the process of incrementally and consistently adapting an existing ontology to changes in the relevant domain. Semantic drift refers to how ontology concepts’ intentions gradually change as the domain evolves. Normally, a semantic drift captures small domain changes that are hard to detect with traditional ontology management tools or ontology learning methods, but may be important to the maintenance of the ontology. This paper discusses a new approach to detecting semantic drift that makes use of concept signatures reflecting the textual references to concepts over time. Comparing how signatures change over time, we see how concepts’ semantic content evolves and how their relationships to other concepts gradually reflect these changes. An experiment with the DNV’s business sector ontology from 2004 and 2008 demonstrates the value of this approach to ontology evolution.

Quality of subsumption hierarchies in ontologies

Ontologies are becoming increasingly more popular tools for many tasks, such as information integration, information retrieval, knowledge management and extraction etc. The cost and complexity of developing good ontologies is high, and therefore it is important to be able to verify the ontology and detect flaws early. In this paper we propose an approach to expose desirable properties of ontological structures. The approach is based on an ontological profile which is an ontology extended with a vector of weighted terms describing the semantics of each concept of the ontology. We describe four hypotheses for the relations among the classes of the ontology and perform experiments to verify them. Our initial findings are that the experiments support the hypotheses.

Contextual search navigation using semantic tag signatures

Search has been and will continue to be an important tool for users who need to locate information in an ever increasing mount of resources. Not all queries have a well defined information need that can easily be described by a keyword query. Exploratory search is one such type of search where the user is not necessarily proficient in the domain or does not have a clear idea of what he is looking for. In such types of search, navigation is beneficial to guide the user in his quest. In this paper we present an approach to contextual navigation search, based on a hierarchical structure constructed from folksonomy tags. The tags are associated with an extended semantic representation used to guide the navigation. Five semantic navigators are introduced, which are navigation strategies the user can benefit from. We present a prototype which has been implemented to show the applicability of the approach to the problem at hand. The preliminary results are promising and demonstrate the ability to direct the user at interesting navigational suggestions and documents.