Imagine you’re the head of a physics department and you want to compare two different research fields: the cosmic microwave background and gravitational waves.
Which receives the most funding? Which is growing fastest? Which is leading to the most technological innovation? There are some information sources available around the web, often free: you can search for specific research proposals on Gateway to Research, and patent office data is also publicly available if you know how to use it. But all of this information is piecemeal, incomplete, and lacking a physics perspective or terminology.
Particularly at a time when the funding landscape is due to undergo a massive overhaul, it’s the IOP’s job to support our members by helping them to understand the landscape, and by advocating for the community with robust evidence about the strengths of physics and its economic impact.
Even seemingly simple questions, such as how much funding went to UK physics research last year, have no definitive answers at the moment – and that’s where Physics 2020 comes in.
The IOP science and innovation team works to support the physics community, enable knowledge exchange, encourage collaborations, and stimulate innovation. To feed in to these aims we are launching Physics2020 to provide robust evidence about physics in the UK. Initially Physics2020 focused on building a clear picture of physics encompassing all the sub-disciplines. We are now working with research councils, the Intellectual Property Office, physics departments and our special interest groups to collate data and build visualisations to identify trends and intelligence relating to the interest of the physics community. Watch this space…
What is physics?
If we want to understand the full extent of physics activity in the UK, we have to start at the beginning and ask a fundamental question: What is physics?
Physics doesn’t fall into perfect taxonomies like you find in animal kingdoms. While a species can only fall into one genus, which in turn falls into one family, a research topic like lasers can fall into at least two wider fields: optics and instrumentation. There are always new terms emerging, and new fields that span multiple disciplines.
To start to get a picture of this complex world, we looked at the words that physicists use every day. These are collected in a thesaurus of physics terms developed by IOP Publishing (IOPP), and are based on keywords in physics journal articles. The thesaurus links narrower terms to broader terms, for example the term lasers is a narrower term to both optics and instrumentation. We can use the thesaurus to map and link physics ideas, and in doing so we start to see central disciplines emerging. Central disciplines are the larger nodes (with more connections) while smaller nodes have fewer connections and tend to be subdisciplines.
This starry night of physics terms, based on the IOPP thesaurus, gives a first glimpse of the landscape, which we plan to enrich with data important to our members, such as funding, economic impact, and collaborations. The underlying thesaurus will play a central role in identifying and categorising physics activity in the UK.
Colours are determined by a clustering algorithm that creates communities based on how many connections they share. These communities show interesting underlying connections in our thesaurus. As you might expect, communities often stay together, but there are interesting exceptions. For example electronics falls into the same community as optics (both blue), and yet electronics sits snugly next to instrumentation. This is because of the complexity of the network of terms, and while they don’t share direct connections they do share sub-disciplines. As a static image it’s not an analytical tool, but rather an artistic demonstration of the disciplines and terms that make up the physics landscape.
Thesaurus: A controlled list of words and terms that can be used to understand and classify a system. It also has structure, providing information about each term’s relationships to other terms. In the world of data a thesaurus lists synonyms as well as narrower, broader, and related terms.
Taxonomy: The hierarchical classification of a system, and far more rigid than a thesaurus. In a taxonomy all terms belong to a single hierarchy, for example your house cat is part of the genus felis, which is part of the family falidea, order carnivore, class mammalia. In a taxonomy your cat can’t be a part of any other hierarchy, but using a thesaurus these rules wouldn’t exist – one species could fall into multiple families. Clearly a thesaurus is not an appropriate way of classifying animals, but it’s a great way of understanding physics.
What do our members need to know to make evidence-based decisions?
Our members continuously shape the UK physics landscape by making decisions about what they should study, research, fund, promote, and develop.
Students shape the landscape when they choose their specialisms, researchers when they design proposals, heads of physics when they lead their teams in new directions, industry physicists when they decide on new approaches to problems or opportunities, and policy advisers when they lobby on behalf of physics in the UK. All of these decisions affect the course of physics in the UK.
But what evidence do these members use to inform their decisions?
Physics 2020 will provide an interactive, online visualisation and query tool, which will seek to show the landscape of physics in the UK (see EPSRC’s Visualising our Portfolio as an example). It will show a complete picture of funding across different research councils, categorised in terms that physicists use and recognise; it will identify areas of excellence, emerging hot topics, and innovations that are likely to have an impact.
To start off with, we’ll focus on research and patent data, but over the course of the project we’ll be able to include richer information provided by members. We’re already working closely with the RCUK and Intellectual Property Office, and have had fruitful meetings with both organisations which are leading to open and collaborative relationships. Both of these organisations spend significant effort collecting and make their data public, so they’re keen to support us in using their data in novel ways.
We’re currently in the process of setting up data connections to automate and sanitise the data collection process. We’ve also begun to perform semantic analysis of RCUK research abstract data – this allows us to identify physics keywords and perform our own rich analysis of the funding data. The semantic analysis is being performed with support from IOPP, who have been crucial partners and continue to share a well of knowledge and resources. Physics 2020 is about forming connections in the data, but also between organisations.