How Does Google’s MUM Search Update Works?
After the recent BERT update, Google is up for the next big update in its search algorithm.
Google has been testing a new technology mainly for the purpose of understanding the complex search queries that people ask in Google’s search.
Im pretty sure even you would have searched again and again for a particular topic and in the process of doing so you woud have wasted alot of your time.
So whats the solution for this? Well Google has been working secretly on this to solve your problem. Read on!
What is MUM Search Update by Google?
Well, the new AI search technology which Google is testing currently is reportedly known as MUM. The Full form of the MUM stands for Multitask Unified Model.
Google’s goal with this new technology is to answer the complex search queries in a simple, efficient way which is helpful for the users.
MUM will be able to understand different languages and formats so that it can deliver the most complete information that the user is looking for in an effective way possible.
MUM can explain complex search queries like a journalist woud do.
Developers, researchers and data scientists at Google keep on improving the AI so that every complex things can be simplified.
As the google mission statement says it makes the information accessible and make it useful for people around the world.
How Would Google’s MUM Search Update Work?
Google says that when people are searching for complex things they take 8 searches on average to get the complete info. Hmm, that’s quite a time-consuming and tedious task for the users.
Let’s say for example, How to plan night out with friends may require you to perform multiple searches in order to consume the complete information while planning an outing with friends.
These kind of queries can be simplified and provided to users in most effective way possible with the help new MUM AI which is being developed and tested by Google.
Google says that in the future it will take only a few searches to get complex things done.
MUM is also built on the transformer architecture which is similar to the BERT algorithm but the important thing here is that MUM can be 1000 times more powerful and beneficial to the users. It can multi-task things.
Google is currently working on MUM to understand information that is presented in text and image format. They are also going to explore the audio and video formats in the future times.
This model is trained accross 75 different languages. For example, lets say you are searching for any information in English but there isnt really great helpful information written in English.
Suppose there is really nice information which is written in Spanish. So in this case Google will gather that information which is written in Spanish and help you provide that information in your preferred language. Pretty cool, right?
Final Thoughts
MUM technology will be simplifying things for better and hustle free user experience while searching on Google.
With MUM, you can probably stop searching multiple queries for a single topic. MUM is still in the experimental mode and the way the search has evolved over time is getting better and better with the implementation of AI.
