Marek Ciglan

From WikiPapers
Jump to: navigation, search

Marek Ciglan is an author.

Publications

Only those publications related to wikis are shown here.
Title Keyword(s) Published in Language DateThis property is a special property in this wiki. Abstract R C
Evaluation of named entity recognition tools on microposts INES 2013 - IEEE 17th International Conference on Intelligent Engineering Systems, Proceedings English 2013 In this paper we evaluate eight well-known Information Extraction (IE) tools on a task of Named Entity Recognition (NER) in microposts. We have chosen six NLP tools and two Wikipedia concept extractors for the evaluation. Our intent was to see how these tools would perform on relatively short texts of microposts. Evaluation dataset has been adopted from the MSM 2013 IE Challenge. This dataset contained manually annotated microposts with classification restricted to four entity types: PER, LOC, ORG and MISC. 0 0
Learning to find interesting connections in Wikipedia Advances in Web Technologies and Applications - Proceedings of the 12th Asia-Pacific Web Conference, APWeb 2010 English 2010 To help users answer the question, what is the relation between (real world) entities or concepts, we might need to go well beyond the borders of traditional information retrieval systems. In this paper, we explore the possibility of exploiting the Wikipedia link graph as a knowledge base for finding interesting connections between two or more given concepts, described by Wikipedia articles. We use a modified Spreading Activation algorithm to identify connections between input concepts. The main challenge in our approach lies in assessing the strength of a relation defined by a link between articles. We propose two approaches for link weighting and evaluate their results with a user evaluation. Our results show a strong correlation between used weighting methods and user preferences; results indicate that the Wikipedia link graph can be used as valuable semantic resource. 0 0
SGDB - Simple graph database optimized for activation spreading computation Lecture Notes in Computer Science English 2010 In this paper, we present SGDB, a graph database with a storage model optimized for computation of Spreading Activation (SA) queries. The primary goal of the system is to minimize the execution time of spreading activation algorithm over large graph structures stored on a persistent media; without pre-loading the whole graph into the memory. We propose a storage model aiming to minimize number of accesses to the storage media during execution of SA and we propose a graph query type for the activation spreading operation. Finally, we present the implementation and its performance characteristics in scope of our pilot application that uses the activation spreading over the Wikipedia link graph. 0 0
WikiPop - Personalized Event Detection System Based on Wikipedia Page View Statistics Wikipedia
Recommendation
Knowledge base
English 2010 In this paper, we describe WikiPop, a system designed to detect significant increase of popularity of topics related to users' interests. We exploit Wikipedia page view statistics to identify concepts with significant increase of the interest from the public. Daily, there are thousands of articles with increased popularity; thus, a personalization is in order to provide the user only with results re- lated to his/her interest. TheWikiPop system allows a user to define a context by stating a set of Wikipedia articles describing topics of interest. The system is then able to search, for the given date, for popular topics related to the user defined context. 0 1
WikiPop - Personalized event detection system based on Wikipedia page view statistics Knowledge base
Recommendation
Wikipedia
International Conference on Information and Knowledge Management, Proceedings English 2010 In this paper, we describe WikiPop service, a system designed to detect significant increase of popularity of topics related to users' interests. We exploit Wikipedia page view statistics to identify concepts with significant increase of the interest from the public. Daily, there are thousands of articles with increased popularity; thus, a personalization is in order to provide the user only with results related to his/her interest. The WikiPop system allows a user to define a context by stating a set of Wikipedia articles describing topics of interest. The system is then able to search, for the given date, for popular topics related to the user defined context. 0 1