SUNNYVALE, Calif. -- (BUSINESS WIRE) -- Silicon Valley based Grokr today announced the beta launch of a mobile search service that allows people to find the information they need without having to search for it. Grokr’s new entity-based service represents the early steps towards a future where search on mobile devices becomes autonomous, predictive and personalized in order to provide people with the information they need to know, just when they need it.
“The decade old, PC-centric, manual and repetitive ask-and-answer model of search just does not work on mobile devices,” said Grokr CEO and co-founder, Srivats Sampath. “We have an ambitious vision to fundamentally re-imagine mobile search, ushering in a future where the knowledge and information that we seek and need just finds us.”
Grokr’s service is comprised of five foundational building blocks – the Grokr Knowledge Graph and Entity Rank, the Grokr User Graph, autonomous and predictive algorithms and the visually immersive Grokr information tiles. Examples of how Grokr works includes:
Grokr Knowledge Graph & Entity Rank
The core foundation of Grokr’s service is the company’s Knowledge Graph and Entity Rank system – a key breakthrough behind creating a new search model for the mobile world. Based on an improved version of Freebase, one of the largest communities of contributed databases about people, places and things, the system Grokr has built allows the service to understand the true meaning of the entity behind keywords and its relation to other entities. The entity-based model was specifically designed for mobile users; based on location and search behavior, the model can return the most desired result that is relevant based on where, when and who the person is.
Grokr’s Knowledge Base has over 700 million facts, 25 million entities on people, places and things and over 2.5 million ranked entities, in addition to rich datasets from numerous other data sources. The Grokr Knowledge Graph allows the service to determine the meaning behind a keyword, interest or location; for words with multiple meanings, the Entity Ranking system will attempt to automatically rank the most appropriate meaning based on users’ location, search history, interests and other attributes. Using a combination of proprietary and advanced machine learning processes, the Grokr Knowledge Graph is constantly evolving, adapting and learning to the changing conditions of the user.
Grokr User Graph
A significant number of searches people do revolve around their intent, location, interests and professions. With this in mind, Grokr created the Grokr User Graph (GUG), which serves as a chart of graphs that represents everything about a person: who they are, what they like, what they read, where they go and where they work. This permission based system uses a person’s searches, likes, interests and other relevant information from services including Facebook, LinkedIn and Twitter, location data from sensors like GPS, accelerometer and magnetometers, and music listening preferences from mobile devices to build one of the most comprehensive profiles about each user. To protect each Grokr user’s personal profile, the data stored in the GUG is anonymized and contains no personally identifiable information. It is this complete understanding of the user, in conjunction with the GUG and predictive algorithms, which creates the capability to fetch and deliver relevant information at just the right time.
Autonomous & Predictive Algorithms
The human brain is capable of remembering only five to seven things at any given time, while the amount of items that people follow online today numbers in the hundreds, if not thousands. Grokr’s powerful machine learning algorithms learn and monitor the things and subjects that are important to people by literally taking over the cognitive functions such as remembering and recall. Using autonomous and predictive algorithms to calculate the half-life of search intent and interests, Grokr constantly monitors ambient signals such as location, time of day and others to continuously deliver relevant information to users.
Grokr Information Tiles
Information in Grokr is displayed within visually stunning tiles and not in the traditional endless sea of blue links. Supported tiles include, traffic, weather, local, news, breaking news, interest, search, event, and movies.
“Our phones know and are able to understand more information about ourselves than our brains will ever be able to process,” said Grokr CTO and co-founder, Nirmal Nair. “Rather than delivering a summary of the Internet’s content, with Grokr, people are no longer required to search through an endless sea of blue links. The user gets an unadulterated, definitive discovery experience.”
Grokr launched its first private beta in October 2012. The company has raised $4.2 million since its founding in 2010 from NEA, U.S. Ventures, Triple Point Ventures and Lerer Ventures.
Grokr is a free iOS service available in the App Store at: https://itunes.apple.com/us/app/grokr/id574612070?ls=1&mt=8
For more information, please visit: http://grokrlabs.com.
Grokr is a search service designed specifically for mobile devices that allows people to find the information they need without having to search for it. Grokr is creating a search platform where the knowledge you seek just finds you. The company has raised $4.2 million since its founding in 2010 from NEA, U.S. Ventures, Triple Point Ventures and Lerer Ventures. The company is based in Sunnyvale, CA.