Ebrahim Bagheri – Publication Page

Evolutionary Fine-Tuning of Automated Semantic Annotation Systems

John Cuzzola, Jelena Jovanovic, Ebrahim Bagheri, Dragan Gasevic
Reference:
John Cuzzola, Jelena Jovanovic, Ebrahim Bagheri, Dragan Gasevic Evolutionary Fine-Tuning of Automated Semantic Annotation Systems. In Expert Systems with Applications, 42 (20): 6864–6877, 2015.
Links to Publication: [www]
Abstract:
Considering the ever-increasing speed at which new textual content is generated, the efficient and effective use of large text corpora requires automated natural language processing and text analysis tools. A subset of such tools, namely automated semantic annotation tools, are capable of interlinking syntactical forms of text with their underlying semantic concepts. The performance of semantic annotation tools depends on the characteristics of the annotation task, primarily characteristics of the text to be annotated, but also some other task-related features, such as the acceptable precision/recall trade-off. Therefore, these tools need to be optimally configured in order to give their best results when applied to different annotation tasks. However, the configuration of semantic annotators is presently a tedious and time-consuming task due to numerous potential configurations. In this paper, we propose an architecture and a genetic algorithm-based method for automating the task of configuring parameter values of any automated semantic annotation tool. This is a novel and unique solution as, to our knowledge, no solution to the problem of configuring semantic annotators has been reported in the literature so far. Our experiments demonstrate our proposed work enables effective configuration of parameters of any semantic annotator system.
Bibtex Entry:
@article {eswa15, author = {John Cuzzola, Jelena Jovanovic, Ebrahim Bagheri, Dragan Gasevic}, title = {Evolutionary Fine-Tuning of Automated Semantic Annotation Systems}, journal = {Expert Systems with Applications}, year = {2015}, volume = {42}, number = {20}, pages = {6864–6877}, url = {http://www.journals.elsevier.com/expert-systems-with-applications}, abstract = {Considering the ever-increasing speed at which new textual content is generated, the efficient and effective use of large text corpora requires automated natural language processing and text analysis tools. A subset of such tools, namely automated semantic annotation tools, are capable of interlinking syntactical forms of text with their underlying semantic concepts. The performance of semantic annotation tools depends on the characteristics of the annotation task, primarily characteristics of the text to be annotated, but also some other task-related features, such as the acceptable precision/recall trade-off. Therefore, these tools need to be optimally configured in order to give their best results when applied to different annotation tasks. However, the configuration of semantic annotators is presently a tedious and time-consuming task due to numerous potential configurations. In this paper, we propose an architecture and a genetic algorithm-based method for automating the task of configuring parameter values of any automated semantic annotation tool. This is a novel and unique solution as, to our knowledge, no solution to the problem of configuring semantic annotators has been reported in the literature so far. Our experiments demonstrate our proposed work enables effective configuration of parameters of any semantic annotator system.} }




Powered by WordPress