
Prof Advaith Siddharthan
Professor Of Computer Science And Society
Biography
Professional biography
I read Physics at the University of Delhi, and Computer Science at the University of Cambridge before gained my PhD in Computational Linguistics at the University of Cambridge (2003). After Postdoctoral research at Columbia University and the University of Cambridge, I took up my first faculty position at the University of Aberdeen in 2009 before joining the Open University’s Knowledge Media Institute in 2017 as a Reader. I have been Professor of Computer Science and society since 2022. I have over 80 peer-reviewed publications and have been PI on grants from UKRI, EPSRC, NERC and ESRC and Co-I on grants from EPSRC, ESRC, H2020, National Geographic and NERC. I currently work on several research projects developing technologies for biodiversity citizen science, including to make these relevant and accessible to primary and secondary schools, while continuing to dabble in disparate topics in Computational Linguistics.
Research interests
My research intersects Citizen Science, Artificial Intelligence, Data Science, and Sustainability Education. I develop socially responsible AI technologies that bridge the divide between professional scientists and lay public, facilitate meaningful public engagement with science and foster attitudinal and behavioural change, particularly around biodiversity issues. I am the academic lead for four citizen science projects at the OU:
My current research investigates science learning within such citizen science projects, especially how citizens can learn alongside artificial intelligence from data. More details of my projects can be found at the Citizen Science and Artificial Intelligence group pages or on my Personal Webpages.
Impact and engagement
Current projects integrate citizen science learning around pollinators into school curricula in the UK and Italy, collecting and analysing data while encouraging schools and students to create habitats and act as pollinator advocates in society. This research is referenced in the UKRI public engagement strategy document and was showcased during Bees’ Needs Week 2020, a public engagement event coordinated by DEFRA.
I previously led the development of novel technologies aimed at public engagement with nature conservation schemes, co-created in partnership with leading UK charities, the Royal Society for Protection of Birds, Bumblebee Conservation Trust and Royal Horticultural Society. Two projects, Blogging Birds (redkite.abdn.ac.uk) and BeeWatch (beewatch.abdn.ac.uk) were among the 8 selected to feature in the RCUK impact summary report for its Digital Economy theme “Celebrating Success in the Digital Economy". Blogging Birds demonstrated the ability to generate complex automated data-driven texts, received an EPSRC prize (Telling Tales of Engagement Competition), and resulted in a publication in the prestigious Communications of the ACM (Siddharthan et al., 2019). BeeWatch was an online citizen science initiative that generated valuable bumblebee records across the UK by integrating artificial intelligence and data science into citizen science. Following publication of our results on bumblebee feeding patterns in Nature Scientific Reports in 2020, we are working with the RHS to use citizen science data to improve pollinator-friendly planting lists.
External collaborations
Key research collaborators on our citizen science projects include University of Aberdeen, Swedish University of Agricultural Sciences, University of Edinburgh and Imperial College London. I work with a much larger set of partners, within and outside of academia, including on European projects such as https://cos4cloud-eosc.eu/.
Projects
Feeling the untouchable: Haptic touch experiences for naturalistic learning
Our transformational vision is to bring to life Learning through Touch, in the natural world and through digital haptic interfaces, fostering a ‘naturalistic intelligence’ and creating affective connections between what we learn and their personal and societal relevance. Science education has struggled to utilise touch adequately beyond the early years, as concepts and models become more abstract, but also the arts emphasise to visual and in many contexts discourage touching. We suggest that reclaiming the value of touch can bring significant innovation in the way young people can come to learn, but also how they understand their urban and natural surroundings, and specifically, the interface between the two. We will augment natural touch with new haptic interfaces to smartphone and tablet touchscreens based on vibrotactile actuators, that will allow students to ‘touch the otherwise untouchable’, such as textures from nature at different magnifications, or electric fields and currents to help them construct conceptual models. Observing and understanding qualitative perceptions of scale, density, texture and pattern crosscut topics in science and the arts, and arts-based methods will be integrated with science learning as part of a STEAM (STEM + the Arts) approach to education. We will also use touch to open the way for schools to participate in emergent areas of radical citizen science, which advocate a more bottom-up and inclusive approach, sensitive to local context and allowing imaginative flexibility in responding to the new questions of a changing global environment. Our research objectives are to: 1. Develop the possibilities of science learning through digital touch across the primary and secondary school syllabus in topics where students are expected to develop conceptual models of abstract concepts; e.g. topics involving fields and flows, starting with Electric Circuits, where we will enable students to touch electricity on circuits shown on a screen to help construct conceptual models accounting for current, voltage, and resistance. 2. Co-design touch-centred Citizen Science methods for school grounds that use textures from nature to develop pattern-led thinking about underlying structures of natural artefacts. Touch (in nature and through digital interfaces) will be used as a starting point to stimulate progressive inquiries about how plants and animals respond to a changing environment. Schools will then modify their grounds in support of wildlife based on their inquiries. 3. Co-design interfaces for creating and experiencing patterns and textures in art through touch, including interfaces to feel existing artworks, and a sketching tool that allows students to feel drawn lines and textures as they draw them. We will explore Edinburgh Collections’ archive of fieldwork sketches from Charles Lyell and Charles Darwin, combining geological and evolutionary observations as part of a STEAM approach. 4. Understand the impact of these digital and natural touch interventions on cognition, learning, interest, creativity and socialisation with respect to science learning and artistic exploration, including for students with visual impairments, for whom touch is central to sensing the world, and for creating, interacting with and sharing artistic creations. Through pursuing these objectives, we will probe the affective possibilities of a new sensing science, which turns abstract into concrete and distant into intimate. As part of this, we will revolutionise citizen science practice for school students, initiating intuitive bottom-up scientific inquiries through affective methods that lead to attitudinal change and environmentally friendly behaviours. At the same time, we will also advance computing research by understanding the strengths and weaknesses of haptic digital interfaces in relation to tactile learning with physical props, and developing best practice for designing haptic interfaces for learning.
Publications
Journal Article
To touch is to know’: haptic inquiry for primary school citizen science (2025)
Blogging Birds: Telling informative stories about the lives of birds from telemetric data (2019)
Extractive and Abstractive Sentence Labelling of Sentiment-bearing Topics (2019)
SaferDrive: an NLG-based Behaviour Change Support System for Drivers (2018)
Recognizing cited facts and principles in legal judgements (2017)
Presentation / Conference
CSS: Contrastive Semantic Similarities for Uncertainty Quantification of LLMs (2024)
iSpot & AI: Integrating FASTCAT-Cloud and PI@ntNET-API in the Cos4Cloud framework (2023)
Empirical Optimal Risk to Quantify Model Trustworthiness for Failure Detection (2023)
Consensus building in on-line citizen science (2022)
Confidence-Aware Calibration and Scoring Functions for Curriculum Learning (2022)
Gender equality work in a distance learning institution (2022)
Summarising Historical Text in Modern Languages (2021)
Incorporating Constraints into Matrix Factorization for Clothes Package Recommendation (2018)
Generating Summaries of Sets of Consumer Products: Learning from Experiments (2018)
Understanding how to Explain Package Recommendations in the Clothes Domain (2018)
Matrix Factorization for Package Recommendations (2017)
Should Learning Material's Selection be Adapted to Learning Style and Personality? (2017)
Automatically Labelling Sentiment-Bearing Topics with Descriptive Sentence Labels (2017)
Bumblebee friendly planting recommendations with citizen science data (2017)
Summarising News Stories for Children (2016)
Summarising the points made in online political debates (2016)
Exploring the impact of extroversion on the selection of learning materials (2016)
Scrutable Feature Sets for Stance Classification (2016)
Lexico-syntactic Text Simplification And Compression With Typed Dependencies (2014)
Investigation into Human Preference between Common and Unambiguous Lexical Substitutions (2011)
Corpora for the conceptualisation and zoning of scientific papers (2010)
Complex lexico-syntactic reformulation of sentences using typed dependency representations (2010)
Reformulating discourse connectives for non-expert readers (2010)
Whose idea was this, and why does it matter? Attributing scientific work to citations (2007)
An annotation scheme for citation function (2006)
Automatic Classification of Citation Function (2006)
Generating Referring Expressions in Open Domains (2004)
Syntactic Simplification for Improving Content Selection in Multi-Document Summarization (2004)