¹û¶³Ó°Ôº

Welcome to the virtual recap of ¹û¶³Ó°Ôº's inaugural Festival of Digital Research & Scholarship, hosted by ARC

In June 2024, we were delighted to welcome you all to our first ever Festival of Digital Research & Scholarship!Ìý

A huge thank you to everyone who took the time to showcase their workÌýand all thoseÌýwho attended!ÌýThe day was a huge success, packed with a wide variety ofÌýKeynotes, Presentations, Posters, and Exhibitor Tables, celebrating the vast range of research and scholarshipÌýthe ¹û¶³Ó°Ôº digital community has to offer and connecting like-minded teams and peopleÌýacross ¹û¶³Ó°Ôº

Facilitated by our ARC experts,ÌýWorkshops offered light introductions to and demonstrations of programming languages, research tools, and key programming features.

Throughout the day, we asked all attendees to vote on best poster and best presentation. Head to the virtual 'Featured Presentations' andÌý'Poster Exhibition' to see the winners!

We hope you enjoyed the day and look forward to seeing you all next year!

In the meantime, be sure to for the latest news and events across ¹û¶³Ó°Ôº's digital research space

The Festival in numbers

Programme revisit

Workshops

JULIA: A fresh approach to scientific computing | G02, North West Wing

Morning session: 1130-1300

Afternoon session: 1430-1600

Location: G02, North West Wing

Julia is a modern, dynamic, general-purpose, compiled programming language. It's interactive, can be used in a REPL or notebooks. While designed for generic programming, it is particularly well suited for numericalÌýand scientific computing, including simulating differential equations, numerical optimisation, machine learning and differentiable programming.

In this workshop we will give a light introduction to the Julia programming language, focusing on the features which make it a good fit for research software. Consisting of an initial demonstration of some key features and Ìýthe interactive notebook programming environment Pluto (), experts from ARC will facilitate a hands-on, Julia try-out session using an example scientific computing task.

PYTHON: Tooling recommendations and a package template for Python research software projects | Haldane Room

Morning session: 1130-1300

Afternoon session: 1400-1530

Location: Haldane Room, Main Building

The Python programming language is widely used in digital research and scholarship, due to its ease of use, wide ecosystem of existing libraries and open-source licencing. Python packages are a way to organize Python code into reusable components, and so vital to building sustainable research software. A common dilemma faced by researchers using Python is the paralysis of choice of toolsÌýavailable for developing, testing, documenting and benchmarking Python packages - making it unclear what constitutes a good choiceÌýof tools.

Showcasing a project started by ¹û¶³Ó°Ôº's Centre for Advanced Research Computing, and using our tooling recommendations and a Pyhton cookiecutter template, we will demonstrate how to quickly & easily set-up a Ìýnew Python project while adhering to good researchÌýsoftware development practices - including automated testing, dependency management and documentation out of the box.

CONTENDER: Novel Architecture Machines in ARC | North Cloisters

Morning session: 1130-1300

Afternoon session: 1400-1530

Location: North Cloisters, Main Building

As Moore's Law fails and an increasing variety of machine learing accelerators become available, students and practitionersÌýof machine learning will need to become familiar with these technologies to inform their decisionsÌýabout which platforms to use.

Graphcore are an innovative UK company who make accelerator systems for machine learning. ProjectÌýContender at ¹û¶³Ó°Ôº's Centre For Advanced Research Computing (ARC) has a Graphcore IPU-POD 16 Direct Attach system - programmable using their own extensionsÌýto PyTorch called PopTorch.

Primarily for those who develop machine learning models in PyTorch and who are interested in using different platforms, the workshop will demonstrate how to run a simple PyTorch program on our Graphcore () novel AI accelerator machine. As an introduction to this and to these machines generally, attendees will convert a simple PyTorch machine learning code to PopTorch and run it on our Graphcore machine.

The Sloane Lab Knowledge Base: Exploring Cultural Heritage Collections | G02, North West Wing

Single session: 1345-1430

Location: G02, North West Wing

The evolving landscape of digital research methodologies and technologies offers a unique opportunity for exploring and understanding cultural heritage collections and asking new and different questions of them.

The Sloane Lab Knowledge Base (KB), a homogeneous data environment, uses formal semantics to allow data integration, semantic enrichment, and knowledge discovery across a disparate set of resources. The most fundamental challenge is the provision of a suitable semantic metadata schema for unifying catalogues and building knowledge graphs to facilitate resourceful query and visualisation. Most museum catalogues are object-centric, but the Sloane Lab’s approach is record-centric, meaning Ìýthe record is the central entity that we represent - facilitating the expression of multivocality. The KB is based on Semantic Web standards (RDF/S, OWL) and our data model is built on top of the CIDOC CRM.

Presented via a case study, the workshop will focus on the intersections between cultural heritage and technology, underscoring the critical concept of 'Collections as Data' and the potential of data-driven research methods in enhancing the accessibility, interpretability, discovery and analysis of collections. It will provide a rich insight into the design and development, our modelling choices and the challenges we addressed, including data disparity, integration facilities, inconsistency, uncertainty, and data absences.