New research publications

  • Mixture models as a method for comparative sociality: social networks and demographic change in resident killer whales. Behavioural Ecology and Sociobiology. 2021.

    We use Bayesian mixed models applied to the study of animal social networks. We make discoveries about killer whale social lives and demonstrate the utility of using mixture models to compare social preferences between networks and between species.

  • MN Weiss, DW Franks, DA Giles, S Youngstrom, SK Wasser, KC Balcom, DK Ellifrit, P Domenici, MA Cant, S Ellis, MLK Nielsen, C Grimes, DP Croft. Age and sex influence social interactions, but not associations, within a killer whale pod. Proceedings of the Royal Society of London B: Biological Sciences. 2021.

    We use drones to observe association, synchronous surfacing, and physical contact within a pod of killer whales, and use Bayesian methods to analyse the data. We find that while age and sex did not detectably influence association network structure, both interaction networks showed significant social homophily by age and sex.

  • M. Nielsen, S. Ellis, J. Towers, T. Doniol-Valcroze, D. Franks, M. Cant, M. Weiss, R. Johnstone, K. Balcomb III, D. Ellifrit, D. Croft. A long post-reproductive lifespan is a shared trait among genetically distinct killer whale populations. Ecology and Evolution. 2021.

    We use Bayesian programming to model the survival and reproduction of different ecotypes of killer whales and discover that both resident and transient killer whales go through menopause.

  • Ingram, C, Drachen, A. How Software Practitioners Use Informal Local Meetups to Share Software Engineering Knowledge, In 42nd International Conference on Software Engineering ICSE 2020, Seoul, Korea, May 2020

           Each month in the UK many thousands of software professionals attend technology 'meetups'. However, little is known about what motivates them to give up spare time to participate. We               interviewed meetup group leaders and surveyed participants to produce the world's first detailed study of technology-dedicated meetup communities. Our findings suggest that meetups                 cater to experienced software practitioners, who use meetups for keeping up to date, building local networks and sharing rich tacit knowledge about software engineering in practice.

  • Georgios Kampanos and Siamak F. Shahandashti. "Accept All: The Landscape of Cookie Banners in Greece and the UK." In IFIP International Conference on ICT Systems Security and Privacy Protection, pp. 213-227. Springer, 2021.
    Accept All - The Landscape of Cookie Banners in Greece and the UK (PDF , 407kb) - 
    A comprehensive study of the state-of-the-art in cookie notice practices across more than 14.5 thousand UK websites that reveals significant non-compliance practices and widespread 'dark patterns' nudging users towards privacy-intrusive choices.

  • Ingram, C, Drachen, A. How Software Practitioners Use Informal Local Meetups to Share Software Engineering Knowledge, In 42nd International Conference on Software Engineering ICSE 2020, Seoul, Korea, May 2020. 
    How do software professionals use meetup? (PDF , 1,339kb)

  • Polyglot and Distributed Software Repository Mining with Crossflow. 
    Konstantinos Barmpis, Patrick Neubauer, Jonathan Co, Dimitris Kolovos, Nicholas Matragkas, and Richard F. Paige. 2020. In Proceedings of the 17th International Conference on Mining Software Repositories (MSR '20). Association for Computing Machinery, New York, NY, USA, 374–384.

Crossflow is a distributed stream processing framework we are developing in the group, which facilitates model-driven design and implementation of polyglot (multi-language) data processing pipelines.

  • Towards model-based development of decentralised peer-to-peer data vaults.
    Alfa Yohannis, Alfonso de la Vega, Delaram Kahrobaei, and Dimitris Kolovos. 2020.  In Proceedings of the 23rd ACM/IEEE International Conference on Model Driven Engineering Languages and Systems: Companion Proceedings (MODELS '20). Association for Computing Machinery, New York, NY, USA, Article 54, 1–8. DOI:https://doi.org/10.1145/3417990.3420043 

Vaultage is a toolkit for engineering data vault applications which replicate personal data on user-owned devices and communicate via end-to-end encrypted channels.

  • An Architecture for the Development of Distributed Analytics Based on Polystore Events
    Zolotas A., Barmpis K., Medhat F., Neubauer P., Kolovos D., Paige R.F. (2021) An Architecture for the Development of Distributed Analytics Based on Polystore Events. In: Gadepally V. et al. (eds) Heterogeneous Data Management, Polystores, and Analytics for Healthcare. DMAH 2020, Poly 2020. Lecture Notes in Computer Science, vol 12633. Springer, Cham. https://doi.org/10.1007/978-3-030-71055-2_5 

A framework for intercepting incoming polystore queries and outgoing query results to facilitate analytics and authorization on hybrid big-data polystores.

  • DAX: Data-Driven Audience Experiences in Esports.
    Kokkinakis, A., Demediuk, S. P., Nölle, I., Olarewaju, O., Patra, S., Robertson, J., York, P., Pedrassoli Chitayat, A., Coates, A., Slawson, D., Hughes, P., Hardie, N., Kirman, B., Hook, J. D., Drachen, A., Ursu, M. & Block, F. O. (2020). In Proceedings for ACM International Conference on Interactive Media Experience, IMX 2020 (Barcelona, Spain), Association for Computing Machinery (ACM), 12 p.

Esports (competitive videogames) broadcasts follow a similar structure to traditional sports. However, due to their virtual nature, a large and detailed amount data is available about in-game actions. This provides an opportunity to incorporate data-driven, interactive storytelling into the audience experience.

  • The Performance Index: A New Way to Compare Players.
    Demediuk, S. et al. (2021). In Proceedings of the 2021 MIT Sloan Sports Analytics Conference (Boston, USA).

Measuring the performance of an athlete in team-based sports is inherently challenging, not the least because in a team, athletes have different roles and functions. In this paper, we present a new way of comparing the performance of athletes, which takes your role out of the equation.

  • Analysing Team communication Dynamics in a One-Off Game Encounter. 
    Tan, E., Wade, A., Kokkinakis, A.; Heyes, G.; Demediuk, S. & Drachen, A. (2021). In Proceedings of HICSS 54: Hawaii International Conference on System Sciences. HICSS-54.

    Playing online games brings strict requirements on the ability of players to communicate. In this paper, we look at communication in online teams and how this varies well-functioning and not so well functioning teams.

  • S. Alahmari, T. Yuan, and D. Kudenko, “Reinforcement learning for dialogue game based argumentation,” in Proceedings of the 19th Workshop on Computational Models of Natural Argument at the 14th International Conference on Persuasive Technology (CMNA@PERSUASIVE), Limassol, Cyprus, 2019, pp.29-37.

The paper explores the use reinforcement learning (RL) to enable a computational agent to learn to act as a worthy dialogue participant. The experiments show that the proposed RL agent is able to learn to argue against different baseline agents in order to win and/or win with minimal number of moves. We are currently exploring how an agent learn to produce more coherent dialogue.

  • L. Malmqvist, T. Yuan, P. Nightingale and S. Manandhar, “Determining the Acceptability of Abstract Arguments with Graph Convolutional Networks,” in Proceedings of The Third International Workshop on Systems and Algorithms for Formal Argumentation at the 8th International Conference on Computational Models of Argument (SAFA@COMMA), Italy, 2020, 2672:47-56.

    Given a collection of arguments, what should we believe? The paper explores the use of graphical neural networks (GNN) to determine the acceptability of an argument within the context of a collection of arguments. The proposed method achieves a new state of the art accuracy improving on past studies by 30 percentage points.

  • F.T. Al-Khawaldeh, T. Yuan, and D. Kazakov, “Integrating Stance Detection and Factuality Checking”, International Journal of Advanced Studies in Computers, Science and Engineering, vol 9, no. 3, pp.1-17, 2020

    The paper concerns automated faked news detection, typically, it concerns how to use stance detection (e.g. whether an item of evidence support or against a claim) as an auxiliary task for the main task of factuality checking in the context of multi-task learning. The proposed model outperforms the state of art deep learning method e.g. Multi-Channel Bi-LSTM-CNN with attention, with Perspectrum - a dataset with claims, perspectives and evidence.

 

  • David L. Borchers, Peter Nightingale, Ben C. Stevenson, and Rachel M. Fewster, “A latent capture history model for digital aerial surveys”, Biometrics (in press), 2020.

    Taking surveys of wildlife using an aerial platform is likely to become increasingly common as drones become more capable. In this paper we developed a statistical method to estimate the number of sea mammals (which are not always visible because they dive deep below the surface) using a two-camera aerial survey.

 

  • Xu Zhu, Miguel A. Nacenta, Özgür Akgün, and Peter Nightingale, “How people visually represent discrete constraint problems”, IEEE Transactions on Visualization and Computer Graphics, vol. 26, no. 8, 2020.

    Some kinds of problems and puzzles involve making a set of choices, such as planning out a project so that everything is completed on time without ever exceeding the available resources. We studied how people visualise problems like these on paper as a first step towards developing a visual modelling tool.

 

  • Ewan Davidson, Özgür Akgün, Joan Espasa, and Peter Nightingale, “Effective Encodings of Constraint Programming Models to SMT”, in Proceedings of the 26th International Conference on Principles and Practice of Constraint Programming, pp. 143-159, 2020.

    In this paper we bridge two quite different languages for describing constraint problems such as scheduling, planning, and industrial design, where a set of choices needs to be made to satisfy constraints. We develop efficient translations from a ‘traditional’ constraint modelling language to the language of SMT, which is used by several solving tools.

Contact us

Professor Anders Drachen
anders.drachen@york.ac.uk
+44 (0)1904 325354