When You Just Want to Talk: Natural Language Search
By: Jess Carson
What’s the one thing we like whenever we walk into a store looking to find a product to solve a problem? Personalization. Every person has specific needs for their problems and the one thing everyone can agree on is this: No one likes hitting options 1, 2, 3, or 4 in order to reach an automated response that’s cold and unwelcoming. The online shopping mentality is quickly becoming the same.
Search engines are taking steps to make the transition from question and answer engines to “natural language” engines. In particular, Bing is taking steps to make its search process conversational, even. “Natural language” is an evolving feature of search engines that brings us one step closer to a more “lifelike” search process.
What is it?
In a nutshell, natural language search is utilizing the engine in a way that reflects normal conversation. When you ask questions like, “what is” or “how does”, you’re using natural language. We’ve been conditioned to use search engines focusing on the key terms required to get us to our solution. For example, if you were trying to figure out how old the current U.S. President was, you’d probably search something like, “Obama’s age”. If you were to use natural language (or talk to the search engine like it’s a person), you’d say something more along the lines of, “How old is President Obama?” Give this a try. You’ll get the same answer either way:
Why Do We Care?
The trend is shifting to natural language thanks to tools like Siri and Cortana. Our “personal assistants” are seemingly becoming increasingly intelligent in a way that speaking naturally to them is now second nature. We can even talk to them about our favorite shows:
Or ask when our team is playing next:
So, it’s only natural that our propensity to utilize search engines in the same way is greatly increasing. Another factor is the user’s inclination to search from their mobile devices. People don’t have time to sit at a computer and find what they need. If someone is out with a group of friends and wants to find a place to eat, the person is more likely to whip out their phone and ask it a question than go to a search engine and fiddle around with different keyword combinations in order to find what they need.
How’s it Evolving?
While Google will give you a multitude of extra information that you might look for when searching for the solution to a question (note how the Obama result yielded his background in the Knowledge Graph as well as similar answer to current presidential candidates), what if you want to continue the conversation to answer other questions? What if you could carry on a typed conversation with your search engine that looked and felt like a conversation you’d have with Siri or Cortana (sans witty comebacks and clever punchlines)?
Let’s look at what Bing does:
Just like Google, Bing provides the answer and a Knowledge Graph to support it. But what if I wanted to continue the conversation and figure out how tall she was? But instead of asking, “How tall is Emma Watson”, we’ll ask, “How tall is she?”
Interestingly enough, Microsoft updated Bing Smart Search in order to include a more natural search experience. With it, you can “talk” to a search engine like you were talking to another person. Those at the helm of the engines are figuring out that we, as consumers, just want to keep the conversation moving forward in as little time (and with as little typing) as possible.
But what does it mean?
As consumers push for increasingly seamless search experiences, the engines move closer to employing and updating natural language within their algorithms. Google’ Hummingbird is continually updating to better understand the user’s intent (to keep from having to constantly re-ask the search engine questions when it gets confused) and RankBrain is quickly becoming the pinnacle of machine learning within search engines.
And as “personal assistants” on our mobile devices evolve further with each major software update, the search engines will have to advance in order to keep up with the growing trend. As Google focuses on the user’s need, the move to natural language search (and the continual evolution of it across all search engines) is going to be an important aspect of how the search landscape works.