A few days ago, Google launched a new feature called the "Knowledge Graph". With this new update, Google is getting one step closer to semantic search by attempting to understand the meaning behind the strings of keywords that users search for. Google sums up the Knowledge Graph nicely by saying that it enables users to search for "things", not "strings".
What is Google's Knowledge Graph?
The Knowledge Graph finds relationships between different items including people, places and things. Google has developed a huge library of over 500 million objects and 3.5 billion facts, sourced from Wikipedia and the CIA's World Factbook amongst others.
At present, the feature is only available for the Google.com index but it will soon be rolled out to other countries in the coming weeks.
Below is the official video by Google introducing the new feature:
Benefits for Google's Users
There are three main benefits:
1. Improved relevancy
Google will be able to better determine which item you are specifically searching for. In some instances they will provide users with a box to select the appropriate item which they are looking for. For example, if a user searches for "taj mahal", the user will be able to choose if they want to see results related to the monument in India or to the famous casino in Atlantic City.
2. A quicker search process
Normally, if you searched for "china" in Google, you may have decided to visit the appropriate Wikipedia page which is the second result in the SERPs. However, Google now provides some details within an information box in the SERPs which includes the capital city, dialing code, GDP etc.
3. Suggestions for further research
When searching for "william shakespeare", the information box provided shows a selection of his most notable plays as well as people who are in some way related to or are similar to William Shakespeare. Users can discover Shakespeare plays that they may not have heard of before for example.
SEO Implications : What Changes in the SERPs?
Our opinion is that sites such as Wikipedia may be affected in the long-term as often people are just searching for one fact; providing an entire web page about a certain item means that more information is provided, but this actually makes it less efficient for a user who is just looking for a specific fact.
The Knowledge Graph is primarily focused on providing facts to users. This means that for the vast majority of business and ecommerce sites, there will be minimal/no impact. It will mostly affect users who are in the research stage of the buying process; not necessarily users that websites should target so early on.
As always we will closely monitor how this new development changes the search landscape.
Google is currently focusing on semantic search in order to deliver more relevant results.
Let’s review what Google’s semantic search is and what it's going to change.
Google to Develop Semantic Search
In the months to come, the US search company will use more and more semantic analysis to understand the real meaning of a keyword and improve the relevancy of its search algorithm by providing users with direct results.
According to the Wall Street Journal, Google’s search engine will therefore undergo a major transformation as this change is expected to impact between 10% and 20% of overall search queries.
To implement semantic search, the search giant will rely on its huge data base of 200 million of entities including places, objects, and people thanks to Metaweb, a company specialized in semantic data that Google acquired in 2010.
What is Semantic Search?
Google will integrate the semantic to its technology of search by keyword in order to better differentiate the value of information in a web page.
For example, semantic analysis allows to differentiate keywords with more than one meaning, such as the country 'China' and the ceramic material ‘china’.
The search engine will also be able to reply to direct questions.
Semantic Search Isn’t New at Google
Google actually already uses semantic-search elements by displaying answers to direct questions.
For example, a search for “what is the population of China”, results in displaying a graph and the figures from the World bank (1,338,299,512 in 2010).
Google Wants to Keep the Lead in Search
The leading search company has a market share of over 66% (in search queries) and over 75% (in revenue) in the US and isn’t ready to let anyone grab some shares of his cake.
To maintain its leading position against Microsoft Bing, now Apple Siri on mobile, Google has no choice but innovate to stay ahead of the competition.
After introducing the +1 button and social search results, Google clearly wants to stay ahead of Facebook and Twitter's interest graphs, and therefore use other kinds of signals for both search and ad relevance.
More Relevancy and ... More Money
His social network, Google+, will therefore be a key asset for its semantic search since it allows the company to have information on a network of people, companies, places and things, and so build relationships between them to be used in its search algorithm.
Instead of showing users results or ads that just match the terms in their search, Google will have to match ads to their search based on their meaning, as well as their context, location, +1s, relationships, etc.
By providing users with direct results rather than links to other websites containing typed keywords, Google could manage that users would spend more time on its sites, and therefore earn more revenue from advertising that account for more than 90% of its overall revenue.
Search engine giant Google has acquired semantic search start-up Metaweb.
Metaweb and its Freebase
Founded in 2005, Metaweb Technologies is a Californian-based company that develops Freebase, a free, massive, and collaboratively edited database of cross-linked data, in order to create a system for building the semantic web.
Freebase allows anyone to contribute, structure, search, copy and use data. The database has information on over 12 million things, including movies, books, TV shows, celebrities, locations, companies and more.
Metaweb actually helps site owners, bloggers, and developers make their sites richer. It enables them to easily and intelligently add content from across the Web. In this way, they can access third party content from leading sources like Twitter, Hulu, and many more.
Check out this video presenting Metaweb and its Freebase:
Google to improve search
By acquiring Metaweb, Google clearly wants to improve its own search offerings. Metaweb's Freebase contains millions of tagged data which will help make Google search smarter.
Understanding queries and web pages is necessary to improve search. As you know, the web isn’t only words, it’s information about things in the real world. Google wants to understand the relationships between real-world entities in order to deliver relevant information more quickly.
A few weeks ago, Dassault Systèmes acquired French search engine Exalead also involved in semantic processing.
Dassault Systèmes, a leader in 3D and Product Lifecycle Management (PLM) solutions, acquires Exalead, a French search engine for about €135 million.
Exalead, the 'French Google'
Founded in 2000 by search engine pioneers, Exalead is a global providers of Search Platforms and Search-Based Applications (SBA) for consumer and business users.
Among Exalead’s worldwide clients, there are leading companies such as PriceWaterhouseCooper and Michelin. More than 100 million unique visitors use Exalead’s technology for search each month.
The company is involved in the Quaero project, an European technology consortium, mainly funded by the French government to develop multimedia search tools.
Reshaping the search engine landscape
Exalead has a platform to apply advanced semantic processing to Web-scale data. It brings structure, meaning and accessibility to previously unused or under-utilized data in the disparate information cloud.
Contrary to other search engines such as Google, Bing and Yahoo, Exalead is not largely known from the public yet.
Nevertheless, its thumbnail previews of the target pages along with the results, and its advanced refining on the results page (language, geographic location, file type, categories) seem very interesting tools to me.