Leiki Focus is a semantic software engine designed to understand and create a detailed profile of any piece of text – whether it’s a complex news article or a brief product description – and intelligently match similar items together. Our technology helps publishers, retailers, and advertisers understand their user interests with SmartProfiles, increase advertising income with SmartSegments, and show their audience relevant content across sites with SmartMedia.
Our unique ontology-based technology takes the accuracy and domain of applicability of automatic targeting to new levels. Since 2000, we have been developing the extensive standards-based ontology and the reasoning engine that is the foundation of our content and user profiling. Including hundreds of thousands of concepts and extraction patterns, the software is able to analyse the topics present in textual content and to build a detailed user profile based on actual (rather than stated) user interests.
Years of effort have been invested to make Leiki Focus as easy to use as possible. For most use cases, whether they are content recommendations or user segmentation, integration consists of simply placing the Leiki script on the web page. Content integration happens automatically by our crawler. An extensive API is available for cases where it is required, and for our analysis customers we offer a simplified single method API interface for quick and easy analysis of either text or URLs.
Unique hybrid personalisation approach based on extremely detailed semantic analysis
Contextual meaning layer matching across varied content types and languages.
Individual behavioural interest analysis in real time.
Advanced targeting tools, including smart time limits, flexible filtering options and adjustable result variety.
Significant R&D efforts over 10 years
Real-time recommendations created with optimised profile update and comparison algorithms.
High speed semantic analysis of content with 180k+ topics.
Fast in-memory indexing, produces recommendations instantly.
Real-time analysis of user feedbacks.
Optimised for scalability to large volumes with high availability.
Content can be easily integrated using common feed formats such as RSS and Atom, through URL scraping or via HTTP-based REST interface.
Can be provided either as a hosted service or installed at the customer’s data center.