This week in viral videos, we were watching a robot dancing like a human, a human dancing like a robot, an RKO rampage and a hotchpotch Star Wars remake.
Dublin: 20.10.2014 05.22AM
Hot on the heels of Bing’s social search revamp, Google have released intuitive new search features that put the search for knowledge, not social connections, at the forefront.
While the social graph model, which Facebook’s Open Graph is based on, focuses on people and their connections, likes and interests, Google’s new Knowledge Graph is focused on just that: knowledge. More than just information, Google’s latest tool can match words with what they are related to in the real world.
Basic search involves matching words – which, to a search algorithm, is nothing more than a string of letters – with the same combination of letters found elsewhere on the web. Over the years, Google have enhanced this service to give us the best results possible, making search more intuitive and predicting our needs based on data collated from years of inputting terms into their search bar.
But the Knowledge Graph takes intuitive search to a new dimension, making the search engine capable of understanding the meaning of a word, not just the combination of letters.
To explain this, Danny Sullivan over at Search Engine Land (who has already seen and tested the Knowledge Graph and gives a comprehensive overview in his article) compares two search terms: ‘Matt Groening’ and ‘Frank Lloyd Wright’. Based on data from countless previous searches, the search engine ‘knows’ that people looking up Matt Groening are looking for different information to those looking up Frank Lloyd Wright.
To the right of each set of search results will be a new panel containing input from the Knowledge Graph. When you search for ‘Frank Lloyd Wright’, this panel tells you his date and place of birth, where he went to school, as well as listing some of the buildings he designed. Results for ‘Matt Groening’ appear with a similar panel, only this time, in addition to his birth and education details, we are given a list of books he has written and told the names of his parents and siblings.
Why? Because a significant number of people looking up ‘Matt Groening’ on Google are trying to find out where he got the names for the characters in The Simpsons, and once you see that his sister’s name was Lisa, his mother’s name was Margaret (Marge) and his father’s name was Homer, all becomes clear – gained from simply looking up the cartoonist’s name.
These summaries of relevant content aim to answer most questions related to a search. In his blog post about the new service, Google’s senior vice president of engineering, Amit Singhal, says, “The information we show for Tom Cruise answers 37pc of next queries that people ask about him.” Imagine: with this service add-on, you can now google a location and get information on events happening there. Or you can look up a rollercoaster and find out how long the ride takes. As this service grows and develops, we could save a lot of time with Google answering most of our questions with a single search. Pub quiz cheats will surely be delighted!
Other great new functions that come with an all-powerful and highly knowledgeable search engine help to eliminate incorrect results coming from ambiguous search terms and also introduce an element of discovery and deeper learning with each search.
Singhal uses the example of ‘Taj Mahal’ to demonstrate how the Knowledge Graph can recognise multiple meanings from one term. If you enter this term into the search bar, you will see the usual results related to the world wonder, but on the right-hand panel, below the information summary, you will be offered the chance to view results related to Taj Mahal the blues musician or Taj Mahal the Atlantic City casino.
This is even more useful in the mobile-optimised version, where typing and retyping terms to find the right one can be a pain. To minimise the effort involved in your search on a mobile device, different options for an ambiguous term (like, 'andromeda') are offered to you on an interactive ribbon at the top of the search results, so you can swipe or tap to select what you’re looking for. Making the most of the touch-screen interface, you can also tap or swipe through the information provided by the Knowledge Graph when browsing on a mobile device.
And within this wealth of information you’ll also find related searches that can help expand your knowledge on the searched-for topic. “Some of the most serendipitous discoveries I’ve made using the Knowledge Graph are through the magical ‘People also search for’ feature,” writes Singhal. “One of my favourite books is The White Tiger, the debut novel by Aravind Adiga, which won the prestigious Man Booker Prize. Using the Knowledge Graph, I discovered three other books that had won the same prize and one that won the Pulitzer. I can tell you, this suggestion was spot on!”
Comparisons can be made between Google’s new Knowledge Graph and Wolfram Alpha’s ‘Computational Knowledge’ search engine, which is partnered with Bing and with Apple’s Siri. But Singhal downplays this in speaking to Sullivan, saying that Wolfram is far more computational and their goal is more about effectively using facts in computations, while Google just wants us all to spend less time searching and more time doing what we love. How nice of them!
Like any new service roll-out, the Knowledge Graph won’t be without its detractors. Publishers can be rightly concerned that information parsed from their sites to populate the ‘knowledge panels’ could mean less traffic coming their way. However, Google believes, once people get a thirst for Knowledge they will drink their fill, clicking through the search results to learn more and more and, eventually, leaving the search engine to delve deeper.
This seems more hopeful than Bing’s direct commercial opportunities presented through ‘Snapshots’, but there’s no reason why this service can’t or won’t evolve to meet this challenge. The opportunity is there – the knowledge panels could include direct links to purchase tickets to that venue you’re searching for. Why not?
What’s more, publishers of factual content may have a lot to gain from this spread of knowledge.
As we become more and more reliant on information garnered from a few seconds of search, there is a possibility that errors could fall through the cracks and be presented as a Knowledge Graph fact. To counteract this, Google provide a ‘Report a problem’ link with every snippet of information from the Knowledge Graph. If it’s determined following this alert that the information does indeed warrant a correction, Google will inform the source.
One consistently popular site in search results, Wikipedia, is particularly excited about this function as it means the collaborative ecyclopaedia gains access to a greater level of feedback on its articles and accuracy than ever before, helping to make the site, and its content, even better.
All-in-all, Knowledge Graph sounds like an exciting addition to Google’s search engine and certainly improves the service as a rapid-fire research tool, while encouraging users to dive into an endless pit of knowledge.
Knowledge Graph is being rolled out to US English users first, but will hopefully reach us all before long.