Bing’s Search “Enlightenment” – Satori
The innovation battle between Google and Bing is never ending, as both continuously strive to generate products that will revolutionize the way we search, work and play. Semantic search is one such search innovation and, if we believe the hype, it will turn search engines into personalized solution engines. The evolution of semantic search has the potential to forever quash those frustrating moments of misunderstanding between man and machine, because semantic search strives to apply a searcher’s underlying intent to the verbiage entered in to the search bar. After all, for as savvy as Google has become, to this day, the engine still doesn’t know I’m looking for Happy Days episode #126 “The Fonz is Allergic to Girls” (11/14/78) when I search for “geronimo.”
The purpose of semantic search is to bring the query process one step closer to differentiating and deducing in a way that parallels the human mind. Before semantic revelations, web indexes relied entirely upon words, keywords to be specific, without taking into consideration meaning and the individual searcher’s intent. We’ve discussed Google’s implementation of semantic search, the Knowledge Graph, in the recent past but what is Bing’s response, and will it move us closer to a solution engine? Bing’s semantic search capabilities are powered by Microsoft’s Satori system, and Microsoft feels that Satori is just what the Internet world is looking for (satori is the Buddhist term for “enlightened”). Let’s explore why Satori is special.
Ars Techinca recently provided a step-by-step analysis of semantic search – a three step process involving information extraction, linking and analysis. After raw data is extracted, it is connected with an entity and relationships between individual entities are mapped to later undergo analysis and categorization. This process is actually an extension of Microsoft Research’s Trinity graph database – a distributed, memory-based storage layer. Google and Microsoft’s semantic search programs both use this logic in compiling data from unstructured web information to create a database of relationships between people, places and things.
In developing Satori, Microsoft has signaled that they understand the importance of aligning search and social to interpret entities (queries), so they’ve more widely applied social contexts to Satori’s results than Google’s Knowledge Graph. Bing’s Satori semantic results may display in the main feed but, as an additional semantic step, Bing simultaneously queries Facebook, Twitter, Yelp, blogs etc., for possible connections the searcher may have to experts or peers with similar topical interests. Satori’s social additives differentiate it slightly from Google, whose Knowledge Graph searches reach only into sites like Freebase, Wikipedia, the CIA World Factbook & Zagat. It’s still too early to name a winner for setting the semantic standard, but Microsoft’s understanding of the co-dependent relationship between search and social might give Satori an edge.
As Google and Bing refine & deploy their programs on a larger scale, they will continue to innovate in their delivery of contextual search results. While the Google brand continues to dominate the search sphere from a public perception standpoint, from an innovation standpoint, Bing is gaining ground via Satori’s underlying logic. Semantic search is in its infancy and both competitors’ programs are still tasked with delivering on pressing issues like performance, speed and, more than anything, relevance to the searcher. Unless Google taps the ever-changing social sphere to provide semantic relevance, the Knowledge Graph will only be as good as their third party data is up to date and comprehensive. The crowd-sourced nature of social is Bing’s semantic search trump card… today.