Headup's information discovery platform retrieves, connects and enriches web content. It is based on patent-pending semantic technology that 'understands' and models the relationships between objects on the web.

1. It connects information across disparate data sources to create a unique Knowledge Graph of the web.
2. Using this graph, Headup scans web pages to identify, disambiguate and extract contextually relevant entities mentioned in them.
3. Once entities have been correctly identified and extracted, Headup retrieves relevant related content from disparate and customizable sources such as Yahoo News, YouTube, Twitter, and many others.

• Automatic entity extraction
• Context-based disambiguation
• Automated bridging between disparate data sources

The overall result attained from the combination of these technologies is that Headup automatically enriches topics with a wide range of highly-relevant quality materials, without encumbering site owners and editors with any additional editorial effort. For a demonstration of our entity extraction and contextual disambiguation capabilities please visit our Entity Extraction Playground.

Headup truly understands the relationships between entities on the web. This is a complex concept but is easily demonstrated through examples:

• Music – Headup understands a "band" is an entity that publishes "albums", and that "albums" are entities that are composed of "tracks".
• Film – Similarly Headup knows that a "movie" is an entity that was brought about through the joint labors of a "director" and "actors".
• Business – Headup realizes the relationships between the entities in this realm too, for example Headup knows that a "company" is an entity that is responsible for "products".

At the heart of Headup's technology is a highly-optimized Knowledge Graph that maps out the attributes of the topics it contains, as well as the relationships each has with the others. Currently it identifies, classifies and relates over 100 million topics and objects spanning everything from sports to botany, and religion to travel. However, as we continue to connect the platform to new data sources and APIs, it continuously "learns" to identify new types of entities.

Headup's text analysis relies on context and is therefore capable of disambiguating entities according to the context in which they're mentioned.

For example, based on the context in which the name "John Mack" appears in a given text, Headup is capable of identifying whether the text relates to John Mack the musician, John Mack the CEO of Morgan Stanley, or John Mack the construction company.

Once an entity has been disambiguated correctly, Headup relies on its Knowledge Graph to pull relevant complementary content about it from the best fitting web resources.

Returning to our example, it's obvious the best sources for obtaining information about the musician John Mack are very different from those relating to the construction company of the same name.

Headup applies reasoning to the results it obtains from disparate data sources and combines them in order to optimize the content it provides. For example, when requesting YouTube for videos of songs from a given album, Headup doesn't simply approach Youtube with the album's name. Instead, Headup first queries its own Knowledge Graph to find all the songs in the album, and only then requests YouTube for the best matching clips for these songs. Obviously this guarantees far greater precision in the results obtained.

Headup retrieves information from data sources in real-time, so the information is always up to date. With Headup, users are guaranteed to get the latest content available for any given topic, be it news, images, videos, weather, tweets, etc.

Recognizing the importance of social networks in users' lives, Headup offers social context for all relevant entities.

For example, connected users using Headup to obtain information about a film are provided information about which of their friends are "Fans' of the film, mentioned it in their status updates, saw it (or plan to see it), etc. Based on the user's location and language, Headup also provides relevant show times and links to ticket purchasing.

Keeping users engaged and monetization pose major challenges to online publishers. They're constantly required to:

• Produce, aggregate and edit fresh content to keep users engaged
• Improve monetization by reducing the bounce rate and exit rate
• Backlink to relevant internal pages to increase page views and improve SEO
• Enrich their sites by embedding external content

Headup takes care of all these tasks automatically and effortlessly:

• Automatic entity extraction ensures all the key topics in the publishers' content are addressed
• Context-based disambiguation ensures the retrieved content is relevant
• Customizable quality content sources ensure publishers they're providing their audiences with the quality content they want to promote
• The cross-platform personalized social context guarantees every reader a unique personalized experience that strengthens their relationship with the publisher's brand and ensures repeat visits

Most topic search engines do little more than aggregation, providing results from various content sources and search engines in a single page. Headup, on the other hand, utilizes its unique technology to bring the freshest, most relevant information on every topic.

• Precision and Coverage – Headup's understanding of entities and their properties ensures the information it provides is relevant. Its ever-growing repository of content sources ensures it's complete.
• Social and Personal – Headup adds a personal touch to search results by providing users with their unique social network's view for every topic.
• Contextual Browsing – Understanding entity relationships enables Headup to offer topic-to-topic contextual browsing.
• Timeliness – Headup relies on real-time data sources to ensure users always see up-to-date, relevant information.