Πλοήγηση ανά Συγγραφέας "Kyriakakis, Alexandros"

Τώρα δείχνει 1 - 1 of 1
  • Μικρογραφία εικόνας
    Τεκμήριο
    Ontology-based search engine with speech recognition.
    (Τ.Ε.Ι. Κρήτης, Τεχνολογικών Εφαρμογών (Σ.Τ.Εφ), Τμήμα Μηχανικών Πληροφορικής Τ.Ε., 2019-03-12) Kyriakakis, Alexandros; Κυριακάκης, Αλέξανδρος
    Searching for data throughout the internet is not always an easy task. The reason for that is that the internet is flooded with data and the most common searching techniques often fail to provide the best possible results for the users’ needs. Therefore, researchers have for years been trying to find alternative ways of searching. Through that research, searching techniques based on semantic technologies have emerged. Over the years, search engines based on semantic technologies have gained a lot of popularity for their efficiency, but they are still under development. The end goal of this thesis is to examine the use of semantic technologies for bioinformatics data annotation by employing domain specific ontologies and also to make use of natural language processing methods in order to aid the non-bioinformatics expert users to successfully search for bioinformatics related tools using natural language through either text or a speech recognition system. In addition, we make use of more searching techniques to compare their efficiency and also to ensure that the framework that was created throughout this thesis, is be able to provide the user with the best possible results for his needs. For our experiments, a variety of questions of users searching for bioinformatics related tools from the SEQanswers.com forum were collected, and after having irrelevant words from those questions removed from two expert bioinformaticians, those questions were used as queries in the proposed framework. For the evaluation of the results, they were compared with the proposed tools that were given by forum users as answers for the collected questions. Our results indicate that in the majority, the framework’s results are relevant to the queries, with each searching technique having different precision and response times.