Speaker: Sarvenaz Karimi, Data61 Seminar Date: Tuesday July 4 12:00pm Brief abstract: A large portion of healthcare related data contains text, such as progress notes, reports of adverse drug reactions, clinical biomedical literature, and death certificates. Biomedical text mining has been focusing on extracting information from these resources and transforming them to actionable inference. In this talk, two projects will be covered: (1) information extraction from adverse drug reaction reports, and (2) information retrieval from medical literature to support decision making. For the former, I will introduce our entity extraction and normalisation methods. For the latter, I will discuss our ongoing research in search in medical literature and its continuation in precision medicine within the TREC (Text Retrieval Conference) evaluation framework. Short Bio: Sarvnaz is a senior researcher with expertise in Information Retrieval, Natural Language Processing, and Machine Learning. She joined CSIRO in 2012 where she started working in language and social computing team as well as collaborating with researchers in CSIRO’s eHealth program. She has worked on different projects, such as Australia at Your Service (expertise search), Adverse Drug Reaction Discovery from Patient Reviews, and Deep Learning for Autocoding of Clinical Text and Death Certificates. Before CSIRO, Sarvnaz was a researcher at NICTA (2008-2012) working on search and information retrieval in medical domain (BioTALA project).  Sarvnaz is active in both NLP and IR communities by serving as area-chair, chair, and programme committee member of a number of local and international conferences, including   SIGIR, CIKM, ACL,  ALTA, and ADCS. She has served as reviewer of a number of journals including ACM Computing Surveys, Artificial Intelligence for Medicine, BMC medical informatics and decision making, Language Resources and Evaluation, and Information Processing & Management. She’s also a volunteer at CSIRO’s Scientist and Mathematician in School (SMiS) programme, promoting ICT to the younger generations. Sarvnaz’s formal background is in Computer Science and Software Engineering. She completed a PhD in Natural Language Processing, specifically on machine transliteration in 2008. Her master’s thesis was in Information Retrieval focusing on query expansion using document clustering techniques. Sarvnaz’s bachelor’s degree was in software engineering.