Gurulingappa, Harsha: Mining the Medical and Patent Literature to Support Healthcare and Pharmacovigilance. - Bonn, 2012. - Dissertation, Rheinische Friedrich-Wilhelms-Universität Bonn.
Online-Ausgabe in bonndoc: https://nbn-resolving.org/urn:nbn:de:hbz:5n-28895
@phdthesis{handle:20.500.11811/5324,
urn: https://nbn-resolving.org/urn:nbn:de:hbz:5n-28895,
author = {{Harsha Gurulingappa}},
title = {Mining the Medical and Patent Literature to Support Healthcare and Pharmacovigilance},
school = {Rheinische Friedrich-Wilhelms-Universität Bonn},
year = 2012,
month = jul,

note = {Recent advancements in healthcare practices and the increasing use of information technology in the medical domain has lead to the rapid generation of free-text data in forms of scientific articles, e-health records, patents, and document inventories. This has urged the development of sophisticated information retrieval and information extraction technologies. A fundamental requirement for the automatic processing of biomedical text is the identification of information carrying units such as the concepts or named entities. In this context, this work focuses on the identification of medical disorders (such as diseases and adverse effects) which denote an important category of concepts in the medical text. Two methodologies were investigated in this regard and they are dictionary-based and machine learning-based approaches. Futhermore, the capabilities of the concept recognition techniques were systematically exploited to build a semantic search platform for the retrieval of e-health records and patents. The system facilitates conventional text search as well as semantic and ontological searches. Performance of the adapted retrieval platform for e-health records and patents was evaluated within open assessment challenges (i.e. TRECMED and TRECCHEM respectively) wherein the system was best rated in comparison to several other competing information retrieval platforms. Finally, from the medico-pharma perspective, a strategy for the identification of adverse drug events from medical case reports was developed. Qualitative evaluation as well as an expert validation of the developed system's performance showed robust results. In conclusion, this thesis presents approaches for efficient information retrieval and information extraction from various biomedical literature sources in the support of healthcare and pharmacovigilance. The applied strategies have potential to enhance the literature-searches performed by biomedical, healthcare, and patent professionals. The applied strategies have potential to enhance the literature-searches performed by biomedical, healthcare, and patent professionals. This can promote the literature-based knowledge discovery, improve the safety and effectiveness of medical practices, and drive the research and development in medical and healthcare arena.},
url = {https://hdl.handle.net/20.500.11811/5324}
}

Die folgenden Nutzungsbestimmungen sind mit dieser Ressource verbunden:

InCopyright