Skip to main content

Google's Hummingbird algorithm for better search

Google yesterday launched an update to it's search algorithm and calls it "Hummingbird". It seems like an add-on for the Knowledge graph search feature which is the Google's initiative towards making search more meaningful.

Before I jump on what Hummingbird feature will add-on to the search capability in Google, let me throw some light on how conventional search work's and what changes knowledge graph feature introduced. Then we will try to understand where Google is trying to go with Hummingbird algorithm by comparing it with the bird Hummingbird!

How search works?  

  • Everyone is creating web pages whether its businesses or  individuals. There are trillions i.e. 10^18 of web pages.As per Google it's 60 trillions and it's continuously increasing. There are around 400,000,000 domains. 150,000 new url's every day. 
  • People can't remember the url's as there are so many for everything, so people use search engines like Google, Yahoo, Bing, Ask etc... to search the content. 
  • Businesses or individuals register themselves to the search engines
  • Search engines navigates the web or all the registered sites, which is called crawling.
  • Crawler crawls the registered page and from there it follows links from page to page.
  • Search engines sorts the pages by their content and various other factors and keep track of it in the index.
  • Now as you search any keyword on the search page, Google runs it's best algorithms behind the scenes i.e. auto complete, synonyms, spelling, query understanding and search methods etc..., looking for clues to better understand what you really mean. 
  • Based on the clues, it fetches the document from the index.
  • Google ranks the results based on various factors, few significant one are: 
1. site and page quality   2. freshness   3. safesearch i.e. reduces the amount of adult web pages, images and videos   4. user context i.e. geography, user history, etc...   5. translation and internationalization   6. universal search
  • And then Google fetches the results in fraction of seconds.


What is Knowledge Graph?
It's a feature embedded into search in 2012 to enhance its search engines results with semantic search information gathered from a wide variety of sources. Basically it's a Google enhancement for more effective user context, for example, if you search for 'Taj Mahal' in the Google, based on user context it will show you 
  1. results for the monument 'Taj Mahal',
  2. all the details about Taj Mahal in the right side of the page, 
  3. details about what other things people searched for when they searched Taj Mahal
  4. it will show other thing's with Taj Mahal i.e. the singer in US.
So, what Google does is based on the keyword and the user context i.e. geography and other things i.e. if it's a noun, if it's a place name, etc.., it searched for all the details about the TajMahal and renders those  details along with the results, which it calls Knowledge Graph.
What is Hummingbird? 
Prior to Google's new algorithm, Hummingbirds are part of the Trochilidae family found only in the Americas. They are unknown in the Eastern Hemisphere.Most astonishing quality of hummingbirds is their ability to broadcast color. Like soap bubbles, hummingbird's color comes from iridescence. It winks on and off, depending on the light source and the angle of the viewer. A Hummingbird's flight speed can average 25-30 mph, and can dive up to 60 mph. Hummingbirds may visit 1,000 flower per day in the quest of food.

Not sure what inspired Team Google to name there update hummingbird, whether it's humming birds speed of winking or there quest for nectar or the color radiance. As per Google's statement it the precision and speed!

How Hummingbird is going to help Google?
As per Google it will help them in conversational search. People, when speaking searches, may find it more useful to have a conversation. 

“What’s the closest place to buy the latest Samsung Galaxy phone near my office?” A traditional search engine might focus on finding matches for words — finding a page that says “buy” and “Samsung galaxy,” for example.

Hummingbird may better understand the actual location of your office, if you’ve shared that with Google. It might understand that “place” means you want some store. It might get that “Samsung galaxy” is a particular type of electronic device carried by certain stores. Knowing all these meanings may help Google go beyond just finding pages with matching words.

As per Google, Hummingbird will give more attention to each word in a query, ensuring that the whole query — the whole sentence or conversation or meaning — is taken into account, rather than particular words. The goal is that pages matching the meaning do better, rather than pages matching just a few words.

What users will get?
Definitly better and relevant results as per the user and query context.

What authors will get?
Content authors will be happy i.e. PageRank algorithm as it now come down to place from where search is getting initiated and more of context oriented.

Who are the non-beneficiary?
Malwares or people who use un-ethical means to rank their pages. Now everyone have to write high quality of content.

Who will be the beneficiary?
End users and Google!
 

Comments

Popular posts from this blog

Pivotal Cloud Foundry Developer Certification - Logging, Scaling and High Availability

 How do you access application logs? cf logs APP_NAME cf start APP_NAME To see the logs of particular pcf sub system. cf logs APP_NAME | grep "API\|CELL" To exclude particular logs cf logs APP_NAME | grep -v "API\|CELL" To see application events i.e. start, stop, crash etc... cf events APP_NAME To display all the lines in the Loggregator buffer cf logs APP_NAME --recent  What are the components of the Loggregator system? Loggregator is the next generation system for aggregating and streaming logs and metrics from all of the user apps and system components in a Cloud Foundry deployment. Primary use: 1. Tail/dump logs using CLI.  2. Stream to 3rd party log archive and analysis service 3. Operators and admins can access Loggregator Firehouse, the combined stream from all the apps and metrics data. 4. Operators can deploy nozzle to the firehouse.  A nozzle is a component that monitors the Firehose for specified events and metrics,

Kumaoni Song/Poem - Uttarakhand meri matrebhoomi

O Bhumi Teri Jai Jaikaara Myar Himaala O Bhumi Teri Jai Jaikaara Myar Himaala Khwar main koot tyaro hyu jhalako-2 Chhalaki kaali Gangai ki dhaara myara Himaala Himaala kaali Gangai ki dhaara myar Himaala Uttarakhand meri matrebhoomi Matrabhoomi ya meri pitrabhoomi O Bhoomi teri jai jai kaara myar Himaala Himaala teri jai jai kaara myar Himaala Tali tali taraai kuni-2 O kuni mali mali bhabara myar Himaala Himaala Mali mali bhabara myar Himaala Badari Kedara ka dwar chhana-2 Myara kankhal Hariwara myar Himaala Himaala kankhal Haridwara myar Himaala Kaali Dhauli ka bali chhali jaani-2 Bata naan thula kailasha myar himaala  Ho Bata naan thula kailasha myar Himaala Parvati ko myaro mait yen chha-2 Ho yen chha Shivjyu ko saurasa myar Himaala Himaala Shiv jyu ko saurasa myar Himaala Dhan mayedi mero yo janama-2 Himaala teri kokhi mahana myar Himaala Himaala teri kokhi mahana myar Himaala Mari jula to tari julo-2 O eju ail tyara baana myar Himaala-2 Himaala ail tyara

OpenStack - Conceptual architecture showing the relationship b/w services

AWS vs Openstack comparison https://redhatstackblog.redhat.com/2015/05/13/public-vs-private-amazon-compared-to-openstack/