Google Maps Simplification with Natural Language Processing (UX Case Study)
Natural language processing (NLP) allows software to interpret, analyze and create data points from natural human language.
From a users perspective this is a very efficient and easy way to input data, it requires less thinking, less tapping or clicking and less disruptions in a flow of thoughts. Think about how naturally typing a message to somebody feels, it’s automatic. navigating through a digital product involves taking action on an array of different buttons, toggles selection fields etc. The latter takes more mental processing.
The inspiration for expanding on this idea comes from Todoist, which is a popular to do application. Todoist utilizes NLP when entering tasks, I can type, for example, “garbage out every thursday at 7pm” Todoist will extrapolate the important points from that statement and create a recurring task to take the garbage out and will remind you every Thursday at 7pm. It makes it dead simple to dump tasks from your brain and avoid forgetting them.
I was thinking about other applications for NLP and had the idea that it might work in google maps, specifically route planning involving an “arrive by” or “depart by” parameter because I knew this process involved a lot of steps.
Goal
Shorten the time spent setting up an arrive/depart by route on google maps.
I chatted with a few friends to see if this was a feature in the app that they used often or at all and discovered that the main reasons for using this feature were to:
- Get to class on time
- Get to a job interview early
- Determine correct timing for trips later in the day
Current Flow
Currently, to create an “arrive by trip” it takes roughly seven steps:
- Tap search bar
- Enter destination or tap recent
- Request “directions” on place page
- Select method of transport
- In the case of transit, select preferred route
- Select depart or arrival time
- Start
Build better voice apps. Get more articles & interviews from voice technology experts at voicetechpodcast.com
This can be reduced into a model that looks something like this:
This creates a lengthy linear process, my thinking was to utilize NLP and design a flow that gets the user to the start page quickly and then allows for ad hoc trip refinements.
Flow Using NLP
Firstly, to familiarize people with a new feature I’ve added an inline cue to provide simple instructions. I chose to use this style of onboarding rather than a full page style as its a bit less intrusive.
Search
In the search bar, information is highlighted to alert the user that these bits of information are recognized and being used. This is a way to provide feedback to the user.
Also new would be a notification toggle that could be enabled when users are planning a trip later in the day to alert them that their departure time is coming up.
After timing myself and others creating an arrive by route via the currently available method vs the NLP method (inputing a request i.e. “library by bike for 10am” and hitting enter”) I found that, on average the current method takes 16.05s to set up. The same routes via natural language processing took 7.13s. (Of course this technology doesn’t actually exist in google maps but for the sake of the idea I tried to be as accurate as possible.)
Future
In the future I think there will be a growing place for natural language processing. As more and more people become accustomed to using voice interfaces such as Google Home or Alexa, the need to spend so many steps defining and reviewing actions will start to feel archaic. NLP is a sort of bridge between digital and voice interfaces it allows for people to interact with their devices in an uninterrupted and human way.
I’d also like to note that I’m by no means an expert in natural language processing or machine learning and that this is a purely concept based on my understanding of how that technology works and how I think it could be used to improve products.
ALSO! This is the first thing of length that I’ve really ever written and published on the internet, It feels good to put thoughts to keyboard! This will be the first of hopefully many studies, thoughts and projects that I’ll post here in my journey through design and user experience.
