Skip to main content

Runner up prize for SAMTLA at British Library Labs competition

Birkbeck crest

Mark Levene, Dell Zhang and Martyn Harris were awarded the research runner-up prize at the 2017 British Library Labs Competition for their work on Samtla. This annual competition looks at transformative project ideas that utilise the British Library’s digital collections in original and creative ways.  Samtla (Search and Mining Tools for Labelling Archives) “is a web-based domain-specific research environment for digital humanities, which we created in response to a lack of flexible tools for in-depth research of historic texts” says Martyn Harris.  Samtla models the morphology of a language by using a character-based model, which describes the syntax of the language.  This opens up new possibilities for exploring previously unsearchable specialist sources, in many different languages.  Samtla can provide search over the full spectrum of human culture recorded in written form, without the need to develop specialised software for each specific archive or language.

Samtla includes text mining tools that help researchers to discover historic documents through keyword and phrase search, and corpus browsing through metadata and named entity data extracted from the text.  The difference, states Martyn, is that unlike “the most popular general-purpose web search engines like Google and Bing, Samtla is a language-independent framework that features approximate phrase search and document comparison, which makes it flexible to how people search, as well as how authors express themselves through linguistic devices”.  

The Samtla research team is composed of computer scientists, historians, and linguists who have collaborated to provide a universal platform for mining and analysing text corpora that have remained largely inaccessible to date due to a lack of flexiblity in current large-scale search and labelling approaches. Having successfully deployed Samtla over a number of modern and historic collections, including the Aramaic Magic Bowls from Late Antiquity (Southampton University), the Finacial Times Historic Archive (British Library), and the UK Medical Heritage Library (Wellcome Trust), they plan to continue to work closely with researchers to develop further tools that complement existing research methods, whilst helping researchers to access and analyse original sources.

More news about: