Maya Wardeh

Multi-party argument from experience

A framework, PISA, for conducting dialogues to resolve disputes concerning the correct categorisation of particular cases, is described. Unlike previous systems to conduct such dialogues, which have typically involved only two agents, PISA allows any number of agents to take part, facilitating discussion of cases which permit many possible categorizations. A particular feature of the framework is that the agents argue directly from individual repositories of experiences rather than from a previously engineered knowledge base, as is the usual case, and so the knowledge engineering bottleneck is avoided. Argument from experience is enabled by real time data-mining conducted by individual agents to find reasons to support their viewpoints, and critique the arguments of other parties. Multiparty dialogues raise a number of significant issues, necessitating appropriate design choices. The paper describes how these issues were resolved and implemented in PISA, and illustrates the operation of PISA using an example based on a dataset relating to nursery provision. Finally some experiments comparing PISA with other classifiers are reported.

Database of host-pathogen and related species interactions, and their global distribution.

Interactions between species, particularly where one is likely to be a pathogen of the other, as well as the geographical distribution of species, have been systematically extracted from various web-based, free-access sources, and assembled with the accompanying evidence into a single database. The database attempts to answer questions such as what are all the pathogens of a host, and what are all the hosts of a pathogen, what are all the countries where a pathogen was found, and what are all the pathogens found in a country. Two datasets were extracted from the database, focussing on species interactions and species distribution, based on evidence published between 1950–2012. The quality of their evidence was checked and verified against well-known, alternative, datasets of pathogens infecting humans, domestic animals and wild mammals. The presented datasets provide a valuable resource for researchers of infectious diseases of humans and animals, including zoonoses.

PADUA: a protocol for argumentation dialogue using association rules

We describe PADUA, a protocol designed to support two agents debating a classification by offering arguments based on association rules mined from individual datasets. We motivate the style of argumentation supported by PADUA, and describe the protocol. We discuss the strategies and tactics that can be employed by agents participating in a PADUA dialogue. PADUA is applied to a typical problem in the classification of routine claims for a hypothetical welfare benefit. We particularly address the problems that arise from the extensive number of misclassified examples typically found in such domains, where the high error rate is a widely recognised problem. We give examples of the use of PADUA in this domain, and explore in particular the effect of intermediate predicates. We have also done a large scale evaluation designed to test the effectiveness of using PADUA to detect misclassified examples, and to provide a comparison with other classification systems.

Small animal disease surveillance

This is the first UK small animal disease surveillance report from SAVSNET. Future reports will expand to other syndromes and diseases. As data are collected for longer, the estimates of changes in disease burden will become more refined, allowing more targeted local and perhaps national interventions. Anonymised data can be accessed for research purposes by contacting the authors. SAVSNET welcomes feedback on this report.

Using open-access taxonomic and spatial information to create a comprehensive database for the study of Mammalian and avian livestock and pet infections ☆

What are all the species of pathogen that affect our livestock? As 6 out of every 10 human pathogens came from animals, with a good number from livestock and pets, it seems likely that the majority that emerge in the future, and which could threaten or devastate human health, will come from animals. Only 10 years ago, the first comprehensive pathogen list was compiled for humans; we still have no equivalent for animals. Here we describe the creation of a novel pathogen database, and present outputs from the database that demonstrate its value. The ENHanCEd Infectious Diseases database (EID2) is open-access and evidence-based, and it describes the pathogens of humans and animals, their host and vector species, and also their global occurrence. The EID2 systematically collates information on pathogens into a single resource using evidence from the NCBI Taxonomy database, the NCBI Nucleotide database, the NCBI MeSH (Medical Subject Headings) library and PubMed. Information about pathogens is assigned using data-mining of meta-data and semi-automated literature searches. Here we focus on 47 mammalian and avian hosts, including humans and animals commonly used in Europe as food or kept as pets. Currently, the EID2 evidence suggests that: • Within these host species, 793 (30.5%) pathogens were bacteria species, 395 (15.2%) fungi, 705 (27.1%) helminths, 372 (14.3%) protozoa and 332 (12.8%) viruses. • The odds of pathogens being emerging compared to not emerging differed by taxonomic division, and increased when pathogens had greater numbers of host species associated with them, and were zoonotic rather than non-zoonotic. • The odds of pathogens being zoonotic compared to non-zoonotic differed by taxonomic division and also increased when associated with greater host numbers. • The pathogens affecting the greatest number of hosts included: Escherichia coli, Giardia intestinalis, Toxoplasma gondii, Anaplasma phagocytophilum, Cryptosporidium parvum, Rabies virus, Staphylococcus aureus, Neospora caninum and Echinococcus granulosus. • The pathogens of humans and domestic animal hosts are characterised by 4223 interactions between pathogen and host species, with the greatest number found in: humans, sheep/goats, cattle, small mammals, pigs, dogs and equids. • The number of pathogen species varied by European country. The odds of a pathogen being found in Europe compared to the rest of the world differed by taxonomic division, and increased if they were emerging compared to not emerging, or had a larger number of host species associated with them.

Argumentation based tools for policy-making

Citizens have a variety of ways to consult with their representatives about policy proposals, seeking justifications, objecting to all or part of it, or making a counter-proposal. For the first, the representative needs only to state a justification. For the second, the representative would want to understand the objections, which may involve asking some questions. For the third, the citizen would have to provide a well formulated proposal that can then be critiqued from the standpoint of the governments own policy proposal. At the end of such a consultation, users will have aired their proposals, understood the implications, and received feedback on how their proposals contrast to that of the government.

Canine babesiosis and tick activity monitored using companion animal electronic health records in the UK

Recent publications highlighting autochthonous Babesia canis infection in dogs from Essex that have not travelled outside the UK are a powerful reminder of the potential for pathogen emergence in new populations. Here the authors use electronic health data collected from two diagnostic laboratories and a network of 392 veterinary premises to describe canine Babesia cases and levels of Babesia concern from January 2015 to March 2016, and the activity of ticks during December 2015–March 2016. In most areas of the UK, Babesia diagnosis in this population was rare and sporadic. In addition, there was a clear focus of Babesia cases in the affected area in Essex. Until February 2016, analysis of health records indicated only sporadic interest in Babesia largely in animals coming from overseas. Following media coverage in March 2016, there was a spike in owner concern that was geographically dispersed beyond the at-risk area. Tick activity (identified as ticks being removed from animals in veterinary consultations) was consistent but low during the period preceding the infections (<5 ticks/10,000 consultations), but increased in March. This highlights the use of electronic health data to describe rapidly evolving risk and concern that follows the emergence of a pathogen.

PISA — Pooling Information from Several Agents: Multiplayer Argumentation from Experience

In this paper a framework, PISA (Pooling Information from Several Agents), to facilitate multiplayer (three or more protagonists), “argumentation from experience” is described. Multiplayer argumentation is a form of dialogue game involving three or more players. The PISA framework is founded on a two player argumentation framework, PADUA (Protocol for Argumentation Dialogue Using Association Rules), also developed by the authors. One of the main advantages of both PISA and PADUA is that they avoid the resource intensive need to predefine a knowledge base, instead data mining techniques are used to facilitate the provision of “just in time” information. Many of the issues associated with multiplayer dialogue games do not present a significant challenge in the two player game. The main original contributions of this paper are the mechanisms whereby the PISA framework addresses these challenges.

PADUA Protocol: Strategies and Tactics

In this paper we describe an approach to classifying objects in a domain where classifications are uncertain using a novel combination of argumentation and data mining. Classification is the topic of a dialogue game between two agents, based on an argument scheme and critical questions designed for use by agents whose knowledge of the domain comes from data mining. Each agent has its own set of examples which it can mine to find arguments based on association rules for and against a classification of a new instance. These arguments are exchanged in order to classify the instance. We describe the dialogue game, and in particular discuss the strategic considerations which agents can use to select their moves. Different strategies give rise to games with different characteristics, some having the flavour of persuasion dialogues and other deliberation dialogues.

Arguing from experience using multiple groups of agents

A framework to support “Arguing from Experience” using groups of collaborating agents (termed participant agents/players) is described. The framework is an extension of the PISA multi-party arguing from experience framework. The original version of PISA allowed n participants to promote n goals (one each) for a given example. The described extension of PISA allows individuals with the same goals to pool their resources by forming “groups”. The framework is fully described and its effectiveness illustrated using a number of classification scenarios. The main finding is that by using groups more accurate results can be obtained than when agents operate in isolation.