A sequence of values of a stock price or a wind speed are examples of time series. They are a sequence of measurements of the same quantity over time: as such they are the fundamental observation which we use to understand how systems change. Time series thus have central roles across science and industry: from heart rhythms in medicine, through output fluctuations of factories, to the vibrations of molecules in chemistry. Researchers analyse datasets of time series to yield better understanding of observed phenomena and then often use models of these phenomena, be able to make predictions of future events.

At Imperial College London, Dr Nick Jones has been comparing different time series by computing thousands of different characteristics about them. This giant resource allows powerful approaches to classic problems in data miningallowing us to quickly identify the source of new time series data. In complement to this, over the last 7 years Nick and his collaborators have therefore been building a large database of time series from a variety of areas which allows matches to be identified for new time series.

With the College's EPSRC Pathways to Impact funding, Nick has been working to build a user-friendly tool that allows anyone to take some time series data, drag it onto a website, and then have returned to them other data, drawn from interdisciplinary sources, which has a similar structure. This allows us to contextualize our observations in terms of those from other perspectives/disciplines - rather like being able to use the library code of one book to identify other similar books on the same shelf.