: The ability to rapidly detect outbreaks of emerging infectious diseases is a health priority of global health agencies. In this context, event-based surveillance (EBS) systems gather outbreak-related information from heterogeneous data sources, including online news articles. EBS systems, thus, increasingly marshal text-mining methods to alleviate the amount of manual curation of the freely available text. This paper documents the use of datasets obtained through an EBS system, PADI-Web (Platform for Automated extraction of Disease Information from the web), dedicated to digital outbreak detection in animal health. This paper describes the datasets used for improving 3 important tasks related to PADI-Web, i.e., news classification, information extraction and dissemination.

PADI-web (Platform for Automated extraction of animal Disease Information from the web) is a biosurveillance system dedicated to monitoring online news sources for the detection of emerging animal infectious diseases. PADI-web has collected more than 380,000 news articles since 2016. Compared to other existing biosurveillance tools, PADI-web focuses specifically on animal health and has a fully automated pipeline based on machine-learning methods. This paper presents the new functionalities of PADI-web based on the integration of: (i) a new fine-grained classification system, (ii) automatic methods to extract terms and named entities with text-mining approaches, (iii) semantic resources for indexing keywords and (iv) a notification system for end-users. Compared to other biosurveillance tools, PADI-web, which is integrated in the French Platform for Animal Health Surveillance (ESA Platform), offers strong coverage of the animal sector, a multilingual approach, an automated information extraction module and a notification tool configurable according to end-user needs.

The Platform for Automated Extraction of Animal Disease Information from the Web (PADI-web) is a multilingual text mining tool for automatic detection, classification, and extraction of disease outbreak information from online news articles. PADI-web currently monitors the Web for nine animal infectious diseases and eight syndromes in five animal hosts. The classification module is based on a supervised machine learning approach to filter the relevant news with an overall accuracy of 0.94. The classification of relevant news between 5 topic categories (confirmed, suspected or unknown outbreak, preparedness and impact) obtained an overall accuracy of 0.75. In the first six months of its implementation (January–June 2016), PADI-web detected 73% of the outbreaks of African swine fever; 20% of foot-and-mouth disease; 13% of bluetongue, and 62% of highly pathogenic avian influenza. The information extraction module of PADI-web obtained F-scores of 0.80 for locations, 0.85 for dates, 0.95 for diseases, 0.95 for hosts, and 0.85 for case numbers. PADI-web allows complementary disease surveillance in the domain of animal health.

Global animal disease outbreak detection and monitoring rely on official sources, such as intergovernmental organisations, as well as digital media and other unofficial outlets. Manually extracting relevant information from unofficial sources is time-consuming. The Platform for Automated extraction of animal Disease Information from the web (PADI-web) is an automated biosurveillance system devoted to online news source monitoring for the detection of emerging/new animal infectious diseases by the French Epidemic Intelligence System. The tool automatically collects news via customised multilingual queries, classifies them and extracts epidemiological information. We detail each step of the PADI-web pipeline, with a focus on the new user-oriented features.

Introduction La veille en sante animale a pour objectif l’alerte precoce vis-a-vis de dangers sanitaires connus ou emergents. Elle repose sur le recueil, le suivi et l’analyse quotidienne d’infor- mations issues de sources officielles, telles que l’Organisation mondiale de la sante animale (OIE), et de sources non-officielles telles que les medias ou les reseaux sociaux (Hartley et al. (2010)). Plusieurs systemes de biosurveillance, tels que MedISys (Mantero et al. (2011)), GPHIN (Blench (2008)) ou HealthMap (Freifeld et al. (2008)), sont ainsi dedies a l’acqui- sition et a la diffusion de donnees issues de sources informelles. Ces systemes s’interessent a un large eventail de risques sanitaires (maladies infectieuses humaines, animales ou vege- tales, risques environnementaux, etc.), mais aucun d’entre eux n’est specifiquement dedie a la sante animale. De plus, tous reposent sur une moderation humaine a une ou plusieurs etapes de leur processus. Dans ce contexte, nous presentons PADI-web 1 (Platform for Automated extraction of Disease Information from the web), un outil de biosurveillance des medias digi- taux pour la detection de foyers de maladies animales (Arsevska et al. (2018)). PADI-web est integre dans la thematique de Veille sanitaire internationale, au sein de la plateforme d’Epi- demiosurveillance en sante animale 2 (plateforme ESA). Depuis sa premiere version, dediee a la veille de sources en anglais, PADI-web a ete enrichi d’un nouveau classifieur reposant sur de l’apprentissage automatique et integre les documents multilingues.

This dataset contains a set of news articles in English related to animal disease outbreaks, that have been used to evaluate and train the information extraction module of the PADI-web system (http://epia.clermont.inra.fr/vsi). It is composed of 532 articles (in JSON), with information about the article itself (publication date, title, content, url, etc.) as well as processing information related to the information extraction process (candidates for extraction information, correct or incorrect labels for each candidate). The named entity candidates (locations, diseases, hosts, dates, etc.) have been manually labeled in each article.

After its introduction in Turkey in November 2013 and subsequent spread in this country, lumpy skin disease (LSD) was first reported in the western Turkey in May 2015. It was observed in cattle in Greece and reported to the World Organization for Animal Health (OIE) in August 2015. From May 2015 to August 2016, 1,092 outbreaks of lumpy skin disease were reported in cattle from western Turkey and eight Balkan countries: Greece, Bulgaria, The Former Yugoslav Republic of Macedonia, Serbia, Kosovo, and Albania. During this period, the median LSD spread rate was 7.3 km/week. The frequency of outbreaks was highly seasonal, with little or no transmission reported during the winter. Also, the skewed distribution of spread rates suggested two distinct underlying epidemiological processes, associating local and distant spread possibly related to vectors and cattle trade movements, respectively.

Since 2013, the French Animal Health Epidemic Intelligence System (in French: Veille Sanitaire Internationale, VSI) has been monitoring signals of the emergence of new and exotic animal infectious diseases worldwide. Once detected, the VSI team verifies the signals and issues early warning reports to French animal health authorities when potential threats to France are detected. To improve detection of signals from online news sources, we designed the Platform for Automated extraction of Disease Information from the web (PADI-web). PADI-web automatically collects, processes and extracts English-language epidemiological information from Google News. The core component of PADI-web is a combined information extraction (IE) method founded on rule-based systems and data mining techniques. The IE approach allows extraction of key information on diseases, locations, dates, hosts and the number of cases mentioned in the news. We evaluated the combined method for IE on a dataset of 352 disease-related news reports mentioning the diseases involved, locations, dates, hosts and the number of cases. The combined method for IE accurately identified (F-score) 95% of the diseases and hosts, respectively, 85% of the number of cases, 83% of dates and 80% of locations from the disease-related news. We assessed the sensitivity of PADI-web to detect primary outbreaks of four emerging animal infectious diseases notifiable to the World Organisation for Animal Health (OIE). From January to June 2016, PADI-web detected signals for 64% of all primary outbreaks of African swine fever, 53% of avian influenza, 25% of bluetongue and 19% of foot-and-mouth disease. PADI-web timely detected primary outbreaks of avian influenza and foot-and-mouth disease in Asia, i.e. they were detected 8 and 3 days before immediate notification to OIE, respectively.

Timeliness and precision for detection of infectious animal disease outbreaks from the information published on the web is crucial for prevention against their spread. The work in this paper is part of the methodology for monitoring the web that we currently develop for the French epidemic intelligence team in animal health. We focus on the new and exotic infectious animal diseases that occur worldwide and that are of potential threat to the animal health in France.In order to detect relevant information on the web, we present an innovative approach that retrieves documents using queries based on terms automatically extracted from a corpus of relevant documents and validated with a consensus of domain experts (Delphi method). As a decision support tool to domain experts we introduce a new measure for ranking of extracted terms in order to highlight the more relevant terms. To categorise documents retrieved from the web we use Naïve Bayes (NB) and Support Vector Machine (SVM) classifiers.We evaluated our approach on documents on African swine fever (ASF) outbreaks for the period from 2011 to 2014, retrieved from the Google search engine and the PubMed database. From 2400 terms extracted from two corpora of relevant ASF documents, 135 terms were relevant to characterise ASF emergence. The domain experts identified as highly specific to characterise ASF emergence the terms which describe mortality, fever and haemorrhagic clinical signs in Suidae.The new ranking measure correctly ranked the ASF relevant terms until position 161 and fairly until position 227, with areas under ROC curves (AUCs) of 0.802 and 0.709 respectively.Both classifiers were accurate to classify a set of 545 ASF documents (NB of 0.747 and SVM of 0.725) into appropriate categories of relevant (disease outbreak) and irrelevant (economic and general) documents.Our results show that relevant documents can serve as a source of terms to detect infectious animal disease emergence on the web.Our method is generic and can be used both in animal and public health domain.

In a context of intensification of international trade and travels, the transboundary spread of emerging human or animal pathogens represents a growing concern. One of the missions of the national veterinary services is to implement international epidemiological intelligence for a timely and accurate detection of emerging animal infectious diseases (EAID) worldwide, and take early actions to prevent their introduction on the national territory. For this purpose, an efficient use of the information published on the web is essential. The authors present a comprehensive method for identification of relevant associations between terms describing clinical signs and hosts to build queries to monitor the web for early detection of EAID. Using text and web mining approaches, they present statistical measures for automatic selection of relevant associations between terms. In addition, expert elicitation is used to highlight the most relevant terms and associations among those automatically selected. The authors assessed the performance of the combination of the automatic approach and expert elicitation to monitor the web for a list of selected animal pathogens.

La veille en sante animale, et notamment la detection precoce d'emergences au niveau mondial d'agents pathogenes, est l'un des moyens permettant de prevenir l'introduction en France de dangers sanitaires (Paquet et al., 2006). Cet article presente une plateforme dediee a la collecte de donnees (depeches) utiles pour la veille automatique. Le recueil des depeches s'appuie sur des requetes constituees de mots-cles de maladies, d'hotes et de symptomes appliquees a Google News. Une interface Web a ete developpee pour consulter les articles collectes et parametrer le processus de recueil en definissant de nouvelles combinaisons de mots-cles.

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.