Unlike traditional technologies based on keyword and statistics that can only guess the content of a text, Cogito reads and interprets knowledge trapped in unstructured text, finding hidden relationships, trends, and events. It relies on deep linguistic analysis and semantic disambiguation of text to ensure a complete understanding of a text. The technology can be used on files, e-mails, articles, reports, and Web pages.

After developing Cogito, Expert System partnered with Microsoft and integrated the linguistic and semantic technologies into Microsoft Office. The Cogito Categorizer is also integrated to the SharePartXXL Taxonomy Extension for Microsoft SharePointby the SharePartXXL Cogito Connector. In April 2011, the company was awarded a US patent for the Cogito semantic platform.

Products include Cogito Semantic Search, Cogito semantic Advertiser, and Cogito Answers, and Cogito Intelligence Platform. Expert System positions Cogito Semantic Advertiser as an alternative to Google’s AdSense search keyword ad management tool. The company applies semantic technologies to its contextual ad formula, discerning greater meaning from the text in an article to provide more relevant ads. Cogito Answers can be used to improve customer service, combining semantic analysis of sentiment and customer satisfaction monitoring with advanced natural language customer interaction features.

Profitable from the start and with recent growth at a compound annual growth rate of 50%, Expert System has a client list that encompasses a variety of industries. Customers include Vodafone, Eni Group, Pirelli, Telecom Italia, the Italian Ministry of Defense, RIM and CVS Pharmacy. Competitors are Google, Cisco, Flurry, Nuance Communications, and RAMP. Expert System has a strong following in the mobile search space.

Rita Safranek, February 9, 2012

Sponsored by Pandia.com

Founded in 1998, ClearForest was previously an independent software start-up. It was acquired by Reuters in 2007 and is now part of the Markets division of Thomson Reuters. OpenCalais is a strategic initiative from Thomson Reuters, based on ClearForest technology, to support the interoperability of content across the digital landscape.

OpenCalais is free to use in both commercial and non-commercial settings but can only be used on public content. It can process up to 50,000 documents per day (blog posts, news stories, Web pages, etc.) free of charge.  For users needing to process more than that, there is Calais Professional. While it does not keep a copy of the content, it does keep a copy of the metadata it extracts. Offering a de-facto standard for making content interoperable in a fashion that complies with Semantic Web standards ultimately benefits Thomson Reuters, which is then able to track themes, memes and trends on the Web and to potentially do things like link to relevant content that helps provide context to its readers, customers and other constituents.

After releasing a couple of major upgrades – in particular the incorporation of a whole Linked Data ecosystem underneath OpenCalais for companies, geographies, products and a few other things – with little or no adoption and no fundamentally new capabilities being built, the OpenCalais team, headed by Tom Tague, decided to slow down development and let the market for semantic extraction mature. Thomson Reuters believes that there are massive opportunities for OpenCalais in the areas of news, its integration with social media and its utilization as a massive repository of knowledge.

OpenCalais’ early adopters include CBS Interactive / CNET, Huffington Post, Slate, Al Jazeera, “The New Republic,” the White House and more. Customers include: Kodak, Dow Chemical, Eastman Chemical, NASD, EDS, Boeing, US Dept. Air Force, Reuters, Dow Jones, Thomson Financial. Competitors include Eqentia and Evri. . (I would not include Concept Searching or Ontoprise in this short list due to exogenous complexity factors.)

Stephen E Arnold, January 27, 2012

Sponsored by Pandia.com

During a stint as principal investigator for MIT’s Context Interchange, CTO Kon researched and implemented methodologies for enterprise integration of structured and semi-structured data over independently managed and disparate schema databases. He researched, designed, and prototyped integration engines for distributed multi-database query and caching over heterogeneous, distributed, and partially connected databases. As a member of MIT’s Composite Information Systems Laboratory, Kon published on multi-database integration engines and the use of ontology for bridging database schema. With Intellisophic, he has pioneered innovation in the conceptual management of unstructured information and in the integration of structured, semi-structured and unstructured content.

Intellisophic content is machine-developed, leveraging knowledge from respected referenceworks. The taxonomies are unbounded by subject coverage and are cost-effective to create. The taxonomy library covers several million topic areas defined by hundreds of millions of terms. In addition to taxonomic content, the company offers intelligent solutions, such as enterprise search and retrieval, business intelligence, categorization and classification, compliance management, portal infrastructure, social networking, content and knowledge management, electronic discovery, data warehousing, and government intelligence.

Its strategic alliance partners include Mark Logic, DataLever, SchemaLogic, DFI International, and Mosaic, Inc. Competitors Sandpiper, Intellidimension, and HighFleet. The depth and breadth of Intellisophic’s taxonomies, along with its support of the leading text mining, search, and categorization applications, make it a good solution for many industries. (I would not include Concept Searching or Ontoprise in this short list due to exogenous complexity factors.)

Stephen E Arnold, January 26, 2012

Sponsored by Pandia.com

Founded in 1999 by Jean Delahousse and others, Mondeca asserts that it is the leading European provider of technology for the management of advanced knowledge structures: ontologies, thesauri, taxonomies, terminologies, metadata repositories, knowledge bases, and Linked Open Data.

Based in Paris, France, the company has been financed by its founders, as well as investment funds Trinova and Banque Populaire. Before starting Mondeca, Delahousse worked for Andersen Consulting, Paris Stock Exchange and Diagram, a publisher of financial software. With expertise in semantic web, ontologies, and content management, he has experience in the design and launch of large software applications, as well as in implementation of semantic technologies for large international clients.

Mondeca’s products help enterprises to integrate and interlink heterogeneous information by mapping it to explicit knowledge references and improve the way information is retrieved, analyzed, and reused by producing consistent, precise, and relevant metadata as well as supplying the relevant context. Mondeca’s technology is at the core of the Semantic Enterprise Information Architecture that allows to interconnect people and resources as well as to extract the most value from information.

Its products include Content Annotation Manager, a platform for building and managing customized workflows for semantic annotation of content that coordinates content analysis, data mapping, human validation, and knowledge enrichment components; and Intelligent Topic Manager, which supports the management of complex knowledge structures throughout their lifecycle, from authoring to delivery and can be either used independently to store and manage complex domain-specific knowledge structures, or as a service that enhances enterprise search, knowledge discovery, and text mining solutions.

Mondeca has also built its credibility in the Semantic Web space as a key contributor to widely-used international standards: OWL, RDF, SKOS, ISO 25964, and Topic Maps. Clients include Hachette Filipacchi, the World Tourism Organization, and Thomson Scientific. Competitors include Layer2, Wordmap, Data Harmony, and Smart Logic. (I would not include Concept Searching or Ontoprise in this short list due to exogenous complexity factors.)

Stephen E Arnold, January 25, 2012

Sponsored by Pandia.com