Text Mining Methods for Real Time Intelligence on Graphene Enterprise
The project aims to develop new data sources and methods for real-time intelligence to understand and map enterprise development and commercialisation in a rapidly emerging and growing new technology. More specifically, the project focusses on new venture and small and mid-size (SME) enterprise development and commercialisation of graphene. This is a nanoscale two-dimensional material with exceptional properties holding great promise for path-breaking applications across a range of domains including electronics, medicine, batteries, and sensors. The field is expanding rapidly, with thousands of new patents and hundreds of companies already entering the graphene domain.
Project goalsThe project will develop novel and scalable methods to mine and combine information from three sources:
- unstructured enterprise webpages;
- unstructured data from Twitter; and
- data from established structured databases, including data on patenting.
Web pages are used to extract information on enterprise business strategies, trials, tests and new products, funding, managerial and ownership developments, and relationships with other businesses and research organisation. Twitter feeds are accessed and sourced to provide data on fast-breaking developments related to graphene, including developments associated with start-ups and SMEs. Databases on publications and patent applications (such as the Web of Knowledge and Derwent Innovations) are accessed to validate company names and corroborate the presence (or absence) of intellectual property applications and grants by graphene-related topic areas.
Outputs from the information extraction suite are stored in a repository at processing time, so that the information is available on the fly at demonstration time. For instance, users can retrieve graphene based products grown on specific substrates (e.g., epitaxial graphene grows on SiC), properties of graphene (e.g., conductivity, flexibility), which companies produce which products, information about companies, e.g., location, partnerships, funders, social media environments used.
The project is funded by NESTA.
- NaCTeM talk at University of Cambridge mentioned in blog
- Invited Talk at Open Science Fair
- Invited Talk at Text and Data Mining Symposium in Cambridge
- Invited talk at the University of Cambridge
- NaCTeM article amongst top 1% most downloaded PLOS ONE articles in 2016
- BioNLP 2017 programme announced
- NaCTeM paper on descriptive clustering at EACL 2017
- New article describing method and enriched corpus for normalising phenotype mentions in text
- Prof. Tsujii awarded Okawa prize
Other News & Events
- New article describing construction of a biodiversity terminological inventory
- Workshop at Biocuration conference at Stanford University
- Invited talk at Text and Data Mining event in Paris
- Invited talk at HBIO 2016
- Keynote talk at DAMDID 2016