The platform queries 1840 information channels, providing 31 000 articles totaling 15 500 000 words daily, runs search through Google and Bing every 30 min, performs Twitter search within 350 000 tweets per minute. Simultaneously, Graze runs linguistic analysis on these results to provide the user with content directly relevant to their business in real time.

Cross language barriers

Graze analyses multilingual data and feeds the user with semantically relevant content, generated by publishers in various language domains. The live feed is automatically translated by Graze to the user’s preferred language, enabling the user to browse through relevant content generated outside his normal scope of browsing. The platform currently provides content in English, French, German, Russian and Bulgarian, and can be expanded to include other languages.

Contextualise content

The web search returns results that are semantically matched with dedicated profiles of interest. Organised in profiles, the data renders any information about rival products, competitors markets and scientific achievements highly usable and transparent. Graze categorises content in 185 categories to complement the specific profiles.

Indexing content

Graze analyses audio, video and text with linguistic technologies assigning context value to their content. Graze does automatic tagging to preserve as much context as possible unlike simple keyword tagging. Moreover, Graze automatically categorises audio, video and text data assigning them to multiple categories, generating an organisation of data based on similarity of content.

Gaining knowledge

The collected information can be further analysed, explored and turned into a business asset:

• reputation management tools can be run on the content to show a company's standing;
• analytical tools and techniques for competitive intelligence can be run on the content to enhance the building of sales, marketing and communication strategies, discover patterns, generate alerts for risks etc.;
• statistics in Graze show the dynamics of the relations between enterprises and topics over a period of time. Graze recognises more than 400 000 legal entities;
• algorithms for event detection, prediction, decision making can be implemented and run upon request.

Technology

Graze integrates innovative data processing and language technologies: a scalable content crawler, text mining semantic modelling, multilingual clustering and summarisation methods, rule-based and hybrid machine translation. Graze platform constitutes a complete end-to-end Big Data solution that can be configured to support the data processing needs of any organisation.