September 11, 2008 |
Today we have seen the beginning of a new initiative on the Official Google Blog to predict the future of search based on opinions of top 10 Google experts. The series has started with a post by Marissa Mayer, VP, Search Products & User Experience.
Marissa begins with a confession that she is a search addict yet she performs about 20% of the total number of queries she could if search was constantly available. To prove that point she mentions a number of questions she was interested in on Saturday simply because these things were mentioned in conversations and she did not know the answer to them. She uses these questions to demonstrate how search technologies of the future will improve to incorporate solutions to the problems with the questions similar to those she asks.
First of all, search should become easily available everywhere you may need it – on your phone, in your car or in some handheld device we have no idea of today. An absolutely perfect arrangement would be to have a device that could run search queries in the background based on the words it recognizes in conversations. This brings Marissa to voice-based search where a user could communicate with a search engine with his voice instead of a keyboard only. Better yet (my own variation) if it could also read one’s mind to produce results for questions I have not even pronounced aloud.
The second important trend of the future is natural language processing. It has already become evident that search engines should eventually learn to answer real questions instead of making a user extract keywords for the search engine to understand. I believe that when Marissa Mayer tells natural language is one of important factors in the future of search, it must mean that Google will pay lots of attention to it.
We have recently seen Microsoft claiming that it will use semantic search and natural language processing to bring a better search experience to users and compete with Google – which explained the reason for Microsoft to acquire natural language search engine Powerset. It will no doubt be interesting to see how Google acts in this field and if we should expect some acquisition here or in-house development instead.
Next we should expect Google to integrate different types of search terms input to allow visual and audio search for questions or notions where search engine could find answer or relevant facts without users suggesting words to the search engine at all – only images or sounds. And same is true not only for search query but for search results as well: Google will be working to enhance what is already visible in the Universal search so that a user receives exactly the type of media (be it text, audio, image, video) that provides the best answer or solution.
An important part of the search engines of the future will be personalization for search engines to understand more about the user and what this user wants to achieve when performing a search based on information the search engine knows about the user. Here Marissa emphasizes the fact that users will have full control over what personal information they provide but does her best to explain why exactly search engines will need information about the user to provide more relevant search results. I believe the search giant still wants to persuade users that they should care less about their privacy as compared to relevancy but I don’t expect many users to be actually excited about the concept of a search engine as your best friend that knows everything about you and provides you with information based on what it already knows.
Unfortunately, at least one part in this personalization concept scares me – and this is the social component of search. Marissa Mayer suggests that a search engine of the future is supposed to understand the nature of my question better if it knows who my friends are and some background information about me and these people. Honestly, this sounds like too much: I’d really rather spend extra few minutes sending additional queries to Google than knowing that when I search for a Japanese restaurant in a particular district of the city, Google already knows who of my friends I will be meeting in that particular restaurant and that we will most probably be talking about the upcoming wedding of one of the friends because yesterday she spent an hour on Google searching for various wedding gowns. This may not be a bad idea to get results quickly but knowing the search engine already knows everything about me and my friends because we provide all our information willingly sounds like a big brother watching all of us. So no, thank you, I will be happy to provide you with my location information and even with some information on my occupation and interests but I’d prefer to keep my social graph along with my personal information to myself – I will sure need an opt-out here.
And finally, the last focus for Google will be in machine translation so that if information is available anywhere online but not in the language that you send a query with, Google could find this information for you, translate it to the language you will understand and bring it back to you. Sure, it sounds just perfect but I tend to be a terrible skeptic when it comes to machine translation and its results – even when it is the almighty Google working on it. But 10 years is a lot of time so it could eventually change, of course, and only time will tell if these efforts in machine translation help.
In general, this has been a very informative piece of reading to me as it explained exactly where we should see Google going and what we should expect it to provide users with. But there is one thing that I am kind of uncomfortable – and that is one of the questions Marissa mentioned in the beginning of the post: “What’s the name of the largest city in Russia after Moscow and St. Petersburg?” Since I happen to live exactly in this third largest city in Russia, I naturally know the name and was somehow pleased it was an interesting question to Google’s vice president who also happens to be a woman and one of the women in technology I admire the most. What’s more, I even felt flattered and anticipated a glory day for my beloved city of birth. But unfortunately even the best search technologies do not prevent spelling mistakes – and this is how “Novosibirsk” (correct spelling) transformed into “Novobirsk”. It’s Novosibirsk, Marissa – for a “new Siberian city” since the city is in Siberia and is pretty new – only 115 years old. So maybe search engines should also ensure some way for users to remember the results they produce?
Photo of loud speaker by woodleywonderworks used under Creative Commens.
Post updated to correct the city’s age.