Text analytics for news and social media: The development of a cloud-based marketplace for text analytics

Text analysis can be immensely valuable to many businesses, but the barriers to entry are often unnecessarily high. Potential users need to either run tools in-house, which involves significant start-up costs, or use existing software-as-a-service providers, who each define their own application programming interfaces (APIs) with potentially significant work involved in moving from one provider to another. On the other side of the coin, small developers of text analysis tools who want to make them available to the public must define their own interfaces and host their services themselves. The AnnoMarket project was conceived to develop an open marketplace bringing together potential users and suppliers of text analysis tools and data resources. All analysis services on the AnnoMarket platform, whether developed by the project or supplied by third parties, are exposed through a common set of APIs and use common data formats so users can quickly and easily try out different services and switch between them with minimal effort. This paper provides a detailed technical overview of the AnnoMarket platform and some of the example services and data resources developed to showcase the platform and act as an initial seed for the marketplace. This paper is based on a presentation delivered to the ISKO UK meeting ‘Taming the News Beast’ in London in April 2014.