Ivana Donevska

This work addresses the problem of integrating semantic
awareness in Trajectories Data Warehouses (TDW) for symbolic
trajectories, for the purpose of extending their querying
capabilities. We incorporate the notion of Semantic Re-
lation (SR) as part of the formal model of a TDW, which enables
capturing the similarity between different annotations
describing Points of Interest (POI), locations and activities
used in specifying symbolic trajectories. We formally… Show more

This work addresses the problem of integrating semantic
awareness in Trajectories Data Warehouses (TDW) for symbolic
trajectories, for the purpose of extending their querying
capabilities. We incorporate the notion of Semantic Re-
lation (SR) as part of the formal model of a TDW, which enables
capturing the similarity between different annotations
describing Points of Interest (POI), locations and activities
used in specifying symbolic trajectories. We formally define
the inclusion of the relationship between different terms
used as descriptors in symbolic trajectories and present the
Semantic Relatedness in Trajectories Data Warehouse (SRTDW)
model. We introduce the newly enabled (categories
of) queries in the SR-TDW model and illustrate with specific
examples the impacts of the added functionality. Our
experimental observations demonstrate the benefits of the
proposed approaches in terms of both enriched as well as reduced
answer-sets for the common OLAP-based queries, and
the sensitivity in terms of the various measures of semantic
similarity. Show less

This paper presents an approach for semantic word comparison by coupling natural text descriptions with semi-structured knowledge for revealing more precise context information. The goal of the study is to present how popularity of a word’s sense can affect semantic relatedness when two words are compared. Sense disambiguation and frequency of use are also introduced to derive the popularity of a word’s meaning. Relatedness as a metric represents the semantic proximity between two words. The… Show more

This paper presents an approach for semantic word comparison by coupling natural text descriptions with semi-structured knowledge for revealing more precise context information. The goal of the study is to present how popularity of a word’s sense can affect semantic relatedness when two words are compared. Sense disambiguation and frequency of use are also introduced to derive the popularity of a word’s meaning. Relatedness as a metric represents the semantic proximity between two words. The relatedness score is higher between words which have popular meanings and are related to each other. If the words are similar, however not often used, their relatedness score decreases. The experimental observations demonstrate that the proposed methodology of frequency of use coefficients yield improvements when detecting relatedness and popularity between words. Show less

Scientific-research project, participated in model 3D animation, 10.2010-09.2012.