For argumentation to be applied successfully to biology a number of questions must be raised and resolved. It is the aim of this work to explore possible answers to these questions through the development of a solution to the use case issues described in Sections 4.1 to 4.3. This section will document the questions to be investigated. The detailed questions are given below, but first a summary is presented using the following abstract queries:
1. Which form of argumentation is appropriate as a solution to the issues described in Sections 4.1 to 4.3 - and how effective is it?
2. How should argumentation be presented to a biologist so that (s)he can under- stand, and utilise it?
3. What insights have been gained from the work, and how does that inform the future use of argumentation within biology?
The detailed questions are listed and explained in the following text.
The first abstract query asked what form of argumentation is appropriate, and how effective is it? Yet, it failed to consider the possibility that argumentation may not be appropriate at all. Might it be the case that simply generating reasons (ar- guments) for and against each side of a query will be sufficient without the added debate (argumentation)?
Alternatively, this first abstract query can be interpreted as asking what argu- mentation actually is? How shall it be implemented in this domain? What form can the argumentation realistically take? What practical constraints exist? What are the consequences of these constraints?
A number of implementation related questions may be posed under the banner of abstract query 1. For example, which architecture appears appropriate for this situation? In addition, how can this argumentation system be deployed in order that it is used by a range of biologists?
Abstract query 2 relates both to the presentation of arguments and argumentation. Traditionally argumentation theory presents arguments using a graphical representa- tion. Several main forms of graphic presentation are featured commonly in tools
and publications. Which of these is favoured by the biologists? Furthermore, is the graphical form actually preferable to a textual depiction? Moreover, how well do the standard argumentation concepts translate to the biological domain?
The final abstract query considers issues relating to the broader use of argumen- tation within biology - effectively what has been learned during this work? Assuming that this process represents the first step towards the widespread deployment of ar- gumentation in biology, what foundations need to be laid before the common uptake of argumentation can take place?
Chapters 5 to 8 document the work undertaken to resolve the above questions, and may contain implicit answers to the above questions; however, explicit answers will not be provided until the work is analysed in Chapters 9, 10 and 11.
4.7
Summary
This chapter examines the problems of conflicting, incomplete and distributed in- formation that affects two biological databases EMAGE and GXD. Although these databases are situated within the field of gene expression for the developmental mouse, the nature of biology ensures that these issues affect most domains to a degree.
Following on from this, the objectives of this work are stated, before a justification of the application of argumentation is provided. Finally a number of questions to be answered by this work are documented.
Chapter 5
Argumentation in biology - a
conceptual consideration
Argumentation is a multifaceted research discipline that can be implemented in a variety of ways for a diverse range of purposes. Central to all of these approaches is the idea of an argument being fundamental - as a unit of reasoning, communication or modelling.
Biology is a productive discipline that generates a large volume of data each year. Chapter 4 explains why some of this information is inconsistent - furthermore it doc- uments the reasons behind the incompleteness of resources that republish biological information online.
This document focuses on a use case in the world of in situ hybridisation gene expression for the developmental mouse. That domain epitomises the problems of inconsistency and incompleteness commonly found in biology. In the previous chapter the objective of this thesis was summarised as:
. . . to improve the understanding of the applicability of computational ar- gumentation within biology through the study of a particular biological use case involving in situ gene expression for the developmental mouse. Whilst tackling the problems faced in the use case, insights into the future role of computational argumentation in biology shall be sought. (Section 4.4)
world with the domain of argumentation. To do so it is necessary to contemplate the following:
• What is an argument? How are these arguments to be used? • Where do arguments come from?
• What can be argued over? How can arguments be attacked?
This chapter shall deliberate on the questions posed here. The first section (Section 5.1) reconsiders the problem examining how it stipulates the abstract notion of the solution by defining what an argument is in this context. This theme is continued into Section 5.2 where the impression of an argument is explored further alongside the strategy for the argumentation process. Section 5.3 takes a pragmatic stance, considering the data available for use in argumentation, and in what manner it may be applied. Throughout this scrutiny, the notion of conflict is pursued. An example of argumentation in biology occupies Section 5.4. The theme of the chapter is concluded, in Section 5.5, with an outline of the solution and the rest of this document, before a summary of the chapter in Section 5.6.