Monitoring Disease Outbreak Events on the Web Using Text-mining Approach and Domain Expert Knowledge

Monitoring Disease Outbreak Events on the Web Using Text mining Approach and Domain Expert Knowledge Elena Arsevska1,2, Mathieu Roche3,4, Renaud Lancelot1,2, Sylvain Falala1,2, David Chavernac1,2, Pascal Hendrikx5, Barbara Dufour6 1 CIRAD, UMR CMAEE, Montpellier, France 2 INRA, UMR CMAEE, Montpellier, France 3 CIRAD, UMR TETIS, Montpellier, France 4 Montpellier University, CNRS, LIRMM, UMR 5506, Montpellier, France 5 ANSES, UCAS, Maisons-Alfort, France 6 EnvA, Maisons-Alfort, France {elena.arsevska, mathieu.roche, renaud.lancelot, sylvain.falala, david.chavernac}@cirad.fr, pascal.hendrikx@anses.fr, barbara.dufour@vet-alfort.fr Abstract Timeliness and precision for detection of infectious animal disease outbreaks from the information published on the web is crucial for prevention against their spread. We propose a generic method to enrich and extend the use of different expressions as queries in order to improve the acquisition of relevant disease related pages on the web. Our method combines a text mining approach to extract terms from corpora of relevant disease outbreak documents, and domain expert elicitation (Delphi method) to propose expressions and to select relevant combinations between terms obtained with text mining. In this paper we evaluated the performance as queries of a number of expressions obtained with text mining and validated by a domain expert and expressions proposed by a panel of 21 domain experts. We used African swine fever as an infectious animal disease model. The expressions obtained with text mining outperformed as queries the expressions proposed by domain experts. However, domain experts proposed expressions not extracted automatically. Our method is simple to conduct and flexible to adapt to any other animal infectious disease and even in the public health domain. Keywords: digital disease detection, text mining, Delphi method 1. Introduction Emerging infectious disease outbreaks (disease outbreaks) are an incising threat to human and animal population in non-affected countries due to globalisation, movement of passengers and international trade. Traditional disease surveillance systems are based on a structured multilevel health infrastructure, which can lead to delays in reporting of disease outbreaks (Chan et al., 2010). As a complement to the traditional disease surveillance systems, recently cre- ated biosurveillance systems focus on timely detection of disease outbreaks by monitoring the information published on the web. In order to acquire information about disease outbreaks, the common biosurveillance system use terms issued from medical dictionaries or terms proposed by ex- perts (Mantero et al., 2011; Freifeld et al., 2008; Collier et al., 2007). Currently, no precise work presents how the biosurveillance systems identify the terms for monitoring the web. The work presented in this paper fills this missing gap. We investigate the use of domain expert knowledge and a text mining approach to propose expressions in order to im- prove the acquisition of web pages with information about disease outbreaks. We evaluate our approach in the animal health domain and one disease in particular, African swine fever (ASF). 2. Methods The work presented in this paper is part of the methodol- ogy that we currently develop for the French epidemic in- telligence team in animal health (VSI1). The VSI team fo- cuses in monitoring animal disease outbreaks which occur outside …

Content truncated to 500 words