A Dataset for Multi-lingual Epidemiological Event Extraction

Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020), pages 4139–4144 Marseille, 11–16 May 2020 c⃝ European Language Resources Association (ELRA), licensed under CC-BY-NC 4139 A Dataset for Multilingual Epidemiological Event Extraction Stephen Mutuvi♠,♣ Antoine Doucet♠ Ga¨el Lejeune♦ Moses Odeo♣ ♠University of La Rochelle, L3i Laboratory, firstname.lastname@univ-lr.fr ♣Multimedia University of Kenya, smutuvi@mmu.ac.ke ♦Sorbonne University Paris, gael.lejeune@sorbonne-universite.fr Abstract This paper proposes a corpus for the development and evaluation of tools and techniques for identifying emerging infectious disease threats in online news text. The corpus can not only be used for information extraction, but also for other natural language processing (NLP) tasks such as text classification. We make use of articles published on the Program for Monitoring Emerging Diseases (PROMED) platform, which provides current information about outbreaks of infectious diseases globally. Among the key pieces of information present in the articles is the uniform resource locator (URL) to the online news sources where the outbreaks were originally reported. We detail the procedure followed to build the dataset, which includes leveraging the source URLs to retrieve the news reports and subsequently pre-processing the retrieved documents. We also report on experimental results of event extraction on the dataset using the Data Analysis for Information Extraction in any Language(DANIEL) system. DANIEL is a multilingual news surveillance system that leverages unique attributes associated with news reporting to extract events: repetition and saliency. The system has wide geographical and language coverage, including low-resource languages. In addition, we compare different classifi- cation approaches in terms of their ability to differentiate between epidemic-related and unrelated news articles that constitute the corpus. Keywords: Epidemiology, corpus creation, event extraction, classification, multilingual NLP 1. Introduction Web corpora, describing text corpora created from the Web, have recently become popular due to the availability of a vast amount of electronic texts on the Web. The World Wide Web is a valuable source of data, which enables build- ing corpora with wide-ranging attributes such as varying sizes, languages and domains. Such corpora can be ana- lyzed and utilized in key application areas, among which epidemic intelligence. Epidemic intelligence is an integral component of infec- tious disease early-warning mechanisms. It involves the collection, analysis, and dissemination of key information related to disease outbreaks, with the objective of detect- ing outbreaks and providing early warning to public health stakeholders (World Health Organization, 2014). Disease surveillance mechanisms can broadly be classified as either indicator- or event-based surveillance (Huff et al., 2016). Indicator-based surveillance is the conventional type of surveillance systems which rely predominantly on local health practitioners to identify infectious disease outbreaks, where suspected outbreak cases are subjected to laboratory tests for confirmation. A key determinant of the efficiency of such surveillance methods is the underlying health care infrastructure. Poor infrastructure can result in inaccu- rate and irrelevant information being disseminated, with a likelihood of significant time delays in the dissemination of key information relevant to disease outbreaks (Zhou et al., 2011). Inadequate health infrastructure directly often leads to incomplete geographical coverage when reporting epidemics …

Content truncated to 500 words