About
The Open Argumentation Platform (OAPl) pronounced “
Goals
The tools that make up the platform are designed to support a range of argumentcentric activities such as reasoning over argument resources, dialogically-oriented interaction, manual argument analysis, and automated argument analysis. The suite currently comprises seven software tools which can be combined and configured to form a variety of argument pipelines. These tools all aim to have the following:
- A simple but extensible underlying data model.
- Clear extension points for domain specific analysis & representation tasks.
- Tooling to support import from other formats, e.g. AML, AIF, &c.
- An open source canonical implementation.
- Supporting Documentation.
- Liberal (GPL3) licensing.
- A completely open development model including public GIT repository & public issue/bug tracking.
Tools
OAPl currently comprises the following software tools:
- SADFAce is a simple JSON based Argument description format, software library, and supporting tools to enable developers and researchers to describe arguments and to easily reuse their data. The goal is to make it as easy as possible to incorporate argument data into modern software. SADFace has a simple but extensible model that is compatible with AIF and can serialise other formats such as AML. There are both Python 3 and JavaScript implementations of the core SADFace format as well as supporting tools for SADFace document creation, editing, and manipulation. SADFace currently forms the core of OAPl and is the lingua franca that underpins the integration of the other tools.
- ArgDB is the main Argument Database that provides persistence of SADFace documents. This is a CouchDB based datastore which natively stores SADFace JSON documents. A couch-app is used to provide a web search interface. ArgDB is designed to run in either private/local mode or as a public argument data server as part of the argument web.
- MonkeyPuzzle is a browser-based user interface for manual argument analysis. The interface is centered around a resource pane that holds the resource being analysed and a visual workspace in which a graph based argument visualisation is constructed.
- ALIAS is “A Library for Implementing Argumentation Systems” a Python implemtation of a library for working with Dung frameworks. This is currently used to provide a mechanism for automated reasoning over argument resources, for example, from SADFace documents.
- DGDL & ADAMANT are the Dialogue Game Description Language and its associated Python-based dialogue game execution platform. These two technologies work together to enable dialogue games to be specified, run, and managed.
- Canary is the most recent addition to OAPl, an argument mining library that is currently under heavy development. Canary is a Python library that builds on existing natural language toolkits
Citation
Some of the indiivudal tools within OAPl have been published and information about citing them individually can be found on the respective pages of each tool. However, if you want to cite OAPl itself then please use the following:
S. Wells, "The Open Argumentation PLatform (OAPl)" in Computational Models of Argument, Proceedings of the Eigth International Conference on Computational Models of Argument (COMMA 2020). Perugia, Italy.
Support or Contact
Having trouble or want to get involved with OAPl? The easiest way is to just find one of the tools that interests you, fork it's repo, and get hacking. If you're more sociable then feel free to get in touch as well. For the moment, the best person to contact is Simon s.wells@napier.ac.uk