Monkeypuzzle Towards Next Generation, Free & Open-Source, Argument Analysis Tools John Douglas Simon Wells Edinburgh Napier University Edinburgh Napier University Merchiston Campus Merchiston Campus Ediburgh EH10 5DT Ediburgh EH10 5DT john@johndouglas.co s.wells@napier.ac.uk ABSTRACT analysis within more complex workflows, for example De- We introduce a new, free, open-source, web-based argument bategraph4 amongst many others. See [2] for a bibliography analysis tool called Monkeypuzzle. This is designed to provide of argument diagramming tools. Monkeypuzzle has been in- both a foundation for creating and visualising reproducible spired by this rich heritage of past argument analysis tools, argument analyses as well as a flexible framework for in- indeed it’s name is an homage to the common name of the vestigating new and extending existing argument analysis Araucaria tree. Monkeypuzzle adopts those elements that are techniques. both familiar and useful from existing tools, such as the two pane, source text pane and analysis canvas pane, UI pattern CCS CONCEPTS introduced with Araucaria [6]. The specific boxes and arrows visualisation is a variation on the de facto Argument Inter- • Computing methodologies → Discourse, dialogue change Format (AIF) [1] layout found in the OVA/OVA+ and pragmatics; Nonmonotonic, default reasoning and be- tool, utilising circles to depict I -Nodes and diamonds to lief revision; • Information systems → Web interfaces; depict S -Nodes. KEYWORDS 3 MONKEYPUZZLE Argument Analysis, Open Source Tools, Argument Visualisa- Monkeypuzzle is a free, open source, browser-based argument tion analysis tool that has the following features: (1) Complete source-code available under a permissive li- cense - Full source code is available from the ARG@ENU 1 INTRODUCTION GitHub project repository5 under the GPL3 license6 . Monkeypuzzle is a web-based tool, following an open develop- The importance of this is twofold. Primarily, users ment model, with a focus on pure argument analysis, support can build the app into their workflow without risk for flexible deployment, and rapid innovation with respect that it subsequently either becomes unavailable or to both argument analysis and visualisation techniques. A only available under a restrictive or expensive license. range of newer features have been developed that go beyond Secondarily, because the source is available, users the extant tools to address some shortcomings and to sup- can host their own instances and enhance the app port the needs of changing analytical endeavors. The initial to include features that fit their own research goals; feature set has been spurred by ongoing work to develop the Monkeypuzzle thus becomes a platform not only Sustainable Transport Communications Dataset (STCD1 ) for research but also for experimentation with new [9], an e↵ort to develop a large-scale, high quality analysis of argument analysis and visualisation techniques. arguments used within sustainable transport communication (2) Multiple deployment options - The primary mode for behaviour change. During these e↵orts it became appar- of interaction with the app is via the hosted deploy- ent that a modern, free, and open-source argument analysis ment7 however the app is not server dependent and tool was required that could meet the needs of contemporary two o✏ine forms are supported. The app can be run argument analysts, based upon an open development and from a local filesystem by loading the index.html file deployment model that could sustain rapid, demand-driven into a browser. An o✏ine version is also supported innovation. so that the app is cached in the users browser and reloads from there when the user navigates to the 2 RELATED WORK app’s URL, even if the user is o✏ine. There have been a range of argument analysis tools published (3) Simultaneous analysis of multiple source texts - This over the years including Araucaria[6], Rationale2 , Ova/Ova+3 , is the main innovation within the Monkeypuzzle as well as tools that have supported aspects of argument user interface. Multiple source texts, currently set 4 http://debategraph.org 1 5 https://github.com/ADAPT-project/STCD https://github.com/ARG-ENU/monkeypuzzle web 2 6 http://www.reasoninglab.com/ https://www.gnu.org/licenses/gpl-3.0.en.html 3 7 http://ova.arg-tech.org/ http://arg.napier.ac.uk/monkeypuzzle/ 50 18th Workshop on Computational Models of Natural Argument Floris Bex, Floriana Grasso, Nancy Green (eds) 16th July 2017, London, UK to an arbitrary maximum of ten, can be loaded into refined analysis procedures. The authors do not intend to individual tabs on the text panel and a single analysis suggest that the current application is particularly innovative; made within the visualisation panel. This enables a beyond the bringing together of a core selection of proven domain analysis to be created from multiple resources argument analysis techniques in anticipation of a growing something that is difficult to do with other tools. community of developers who might take the app in directions (4) Support for canonical representations of text nodes - contrary to those mapped out in the remainder of this paper. When analysing multiple source texts and attempting to create a single, large domain analysis rather than a 4 CONCLUSIONS & FUTURE WORK series of individual analyses, variations in voice, writ- The full roadmap is detailed online8 and the project is un- ing styles, complexity of language, and completeness der active development. Immediate development goals are of utterance can reduce the coherency and fluidity as follows: to exploit the use of a tested, reliable, and scal- of the resulting dataset. The app supports editing of able Javascript graph layout library, such as d3.js9 or cy- node text into a canonical form whilst also saving toscape.js10 , so that argument graphs can be automatically the original expressions. This enables higher qual- rendered to the screen, minimising the need for users to ity, curated argument datasets to be constructed. manually adjust the placement of nodes. Additionally we This is particularly important as argument research aim to support mapping of selections from disparate source foci move from straightforward argument analyses texts onto the same analysis nodes, e↵ectively merging nodes towards reuse of the resultant datasets, for example, that have the same meaning but di↵erent natural language in natural language generation tools or to support expressions, especially where these have originated from dif- exploration of contentious knowledge domains. ferent resources. The aim here is to support the development (5) Serialisation to a simple JSON format - The needs of of large, high quality, and integrated argument maps and the tool are driving development of a simple, native, corpora across domains rather than being restricted only to JSON-based file format for saving and loading anal- the analysis of a single given source at a time. The resource yses. The aim is to identify new, useful criteria that pane, although currently restricted to textual resources, will can be used to support extension and improvement eventually support analysis of arguments from a variety of of the AIF. Whilst support for the AIF is on the file types, for example, parsing web-pages (HTML), Portable project’s roadmap, it was decided that a more ap- Document Format (PDF), video, and audio files, to enable propriate starting point would be to rapidly account multi-modal argument analysis. for the various kinds of metadata that the STCD Three areas of active research that we are pursuing are, analysis work is uncovering. User research during firstly, the integration of modified versions of storymaps that our development has shown that many researchers incorporate argument structure, secondly, support for e↵ec- who are performing argument analysis desire the tive dialogue analysis, and thridly, support for visualisation at ability to make ad hoc collections of metadata, as scale. Storymaps are Geographic Information Systems (GIS) demanded by their data, and suggest that current that integrate cartographic maps, geospatial data, and nar- tools frustrate this desire. rative driven content. In 2012, ESRI, a developer of GIS and (6) Export to graphics formats - Visualisations can be spatial analytics software, introduced storymaps and went saved for reuse in other contexts using the Portable on to win awards for Best Digital Map Product and Best Network Graphics (PNG) and Scalable Vector Graph- Overall Map Product from the International Map Industry ics (SVG) formats. Association. Storymaps have since been used to good e↵ect (7) Support for hierarchically organised Argumentation in many journalistic contexts and many nice examples can be Schemes - Walton and Macagno propose a hierarchi- viewed at the Storymaps website11 however an area that has cal organisation of Argumentation Schemes [8] which not been exploited is the combination of argumentative data is implemented within the app. This gives structure and metadata with specific locations and journeys so that to the user and aids in the selection of a scheme to arguments can be visualised in the context of the geographic assign to an argument, rather than choosing from locations that they relate to. We believe that this could prove a long list, organised only by scheme set, a user is to be a useful new dimension in the context of how legal able to select a scheme from a range of categories argument, particularly witness testimony, is explored and to drill down to an appropriate scheme. The goal is visualised. Dialogue analysis has not been well supported by to make it easier to select a scheme to characterise the open-source argument analysis tools but the links between an argument by so that more argument analyses argument and dialogue have been recognised for many years, contain comprehensive scheme analyses rather than having been explored by O’Keefe [4] in terms of Argument1 extensive use of the “default” scheme. and Argument2 , or argument as process and argument as product, but also more recently in dialogical extensions to Bootstrapping a new argument analysis tool to this point has 8 https://github.com/ARG-ENU/monkeypuzzle web/issues taken significant e↵ort. Much of the existing work has been 9 https://d3js.org/ preliminary sca↵olding to enable the future implementation, 10 http://js.cytoscape.org/ 11 integration, exploration, and maintenance of both new and https://storymaps.arcgis.com/en/ 2 18th Workshop on Computational Models of Natural Argument 51 Floris Bex, Floriana Grasso, Nancy Green (eds) 16th July 2017, London, UK Figure 1: The default Monkeypuzzle User Interface showing the standard, two-pane UI popularised by Arau- caria. The left-hand pane is the source pane, a tabbed collection of textual resources for analysis. The right- hand pane is the visualisation pane. The source pane can be completely collapsed to give a user more room to freely create an argument diagram independent of any specific source text allowing the app to be used for argument construction and exploration as well as argument analysis. the AIF [7] which operationalises the co-construction of ar- datasets, and to contribute to a healthy and varied eco- gument as a product of dialogue. One approach might be system of argument tools to support further development of to enable dialogues to be annotated according to the rules computational models of argument. of established dialectical games [10] and for the argumen- tative content licensed by the moves within the dialogue, REFERENCES for example statement!challenge!defense sequences, to be [1] C. Chesnevar, J. McGinnis, S. Modgil, I. Rahwan, C. Reed, G. Simari, M. South, G. Vreeswijk, and S. Willmott. 2006. To- extracted into the visualisation. Finally, visualisation at scale wards an Argument Interchange Format. Knowledge Engineering will increasingly become an issue as the sizes of argumen- Review 21, 4 (2006), 293–316. tative datasets and corpora increase. Anecdotally, standard [2] D. Khartabil, S. Wells, and J. Kennedy. 2016. Large Scale Argu- ment Visualization (LSAV). In Proceedings of EUROGRAPHICS box and arrow diagrams often become unwieldy to the point 2016. of unusability at around the 50 to 100 node mark. Yet the [3] M. Lippi and P. Torroni. 2016. Argumentation Mining: State of combined output from increasingly accurate Argument Min- the Art and Emerging Trends. ACM Transactions on Internet Technology (TOIT) 16, 2 (2016). ing tools [3], or the fulfilled promise of the Argument Web [4] D. J. O’Keefe. 1977. Two Concepts of Argument. The Journal [5] will yield argument datasets at a scale where the limits of of the American Forensic Association 13, 3 (1977), 121–128. [5] I. Rahwan, F. Zablith, and C. Reed. 2007. Laying the Foundations current visualisation tools are exceeded. for a World Wide Argument Web. Artificial Intelligence 171 Ultimately we plan for Monkeypuzzle to provide a basis (2007), 897–921. for exploring new argument visualisation techniques, to act [6] C. Reed and G. Rowe. 2001. Araucaria: Software For Puzzles In Argument Diagramming And XML. Technical Report. University as a test-bed for new tools to interact with argumentative Of Dundee. [7] C. Reed, S. Wells, G. W. A. Rowe, and J. Devereux. 2008. AIF+: Dialogue in the Argument Interchange Format. In Proceedings of the 2nd International Conference on Computational Models of 3 52 18th Workshop on Computational Models of Natural Argument Floris Bex, Floriana Grasso, Nancy Green (eds) 16th July 2017, London, UK Argument (COMMA 2008). [8] D. Walton and F. Macagno. 2016. A classification system for argumentation schemes. Argument and Computation 6, 3 (2016), 219–245. [9] S. Wells and K. Pangbourne. 2016. Using Argumentation Within Sustainable Transport Communication. In Argumentation and Reasoned Action, Proceedings of the 1st European Conference on Argumentation, Lisbon 2015, volume 1. College Publications, Chapter 34, 781–801. [10] S. Wells and C. Reed. 2012. A domain specific language for describing diverse systems of dialogue. Journal of Applied Logic 10, 4 (2012), 309–329. 4 18th Workshop on Computational Models of Natural Argument 53 Floris Bex, Floriana Grasso, Nancy Green (eds) 16th July 2017, London, UK