Publication:
Detecting drugs and adverse events from Spanish health social media streams

Thumbnail Image
Identifiers
URI: https://hdl.handle.net/10016/19814
Files
PFC_RevertArenazRicardo.pdf (1.73 MB)
Publication date
2014-05
Defense date
2014-05-30
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Google Scholar
Export
Research Projects
Organizational Units
Journal Issue
Abstract
Adverse Drug Reactions (ADRs) are the 4th cause of death in hospitalized patients. Despite the importance of clinical trials, they have many limitations mainly based on time and population. Therefore, other ways of spotting ADRs had to be created, as for instance, the healthcare professionals reporting systems and the spontaneous patients reporting systems created by the FDA or the EMA. Nevertheless, it has been proven that the results obtained are not yet as satisfactory as expected. Health-related social media can be used along with these reporting systems in order to obtain possible information from a source where patients feel more comfortable sharing their experiences by exchanging information. Therefore, the creation of the first corpus annotated with drugs and adverse events from social media in Spanish in order to train and evaluate machine-learning techniques is one of the main goals throughout this project. Furthermore, the implementation of a dictionary-based approach to detect mentions and to be tested with this gold-standard is the other main goal in this investigation.
Description
Keywords
Tecnología de la información, Minería de datos, Medicamentos, Medios de comunicación
Bibliographic citation
Collections
Proyectos Fin de Carrera