Cloud Natural
Language API
https://cloud.google.com/natural-
language/overview/docs/
When analysing
customer service
costs, the friendly
customer service
person sitting in
front of a
workstation tends to be the most significant- meaning it’s in your interest to maximise
their time. If programs could analyse text,
bots could shave off precious seconds from
every interaction between unhappy
customers and expensive support employees.
Google’s Cloud Natural Language API
is dedicated to doing exactly that. Plug in
a query in the user’s own words into the
product and it returns a field of information
related to all kinds of related metadata
it has found. In the AutoML edition, you
are furthermore allowed to create custom
models that are more tailored than the API’s
standard pre-defined categories, enabling
you to focus on more niche expertise or areas
of knowledge.
Google Cloud Vision API
https://cloud.google.com/vision/
Whether you’re
looking at bringing
in image screening
or want to offer
users functionality
based on their
images, being able to understand user-
generated and user-uploaded images can be
invaluable. Sadly, creating neural networks
by hand is an incredibly tedious job that
takes lots of time and an insane amount of
training images.
Google’s Cloud Vision API lets your
programs tap into Big G’s machine learning
systems. Upload an image or two and feast
your eyes on the vast amounts of image data
the company has at its disposal. Not only can
the Cloud Vision API determine if an image
contains explicit or infringing content, it
can easily identify the content of images and
even highlight specific features – useful ifCloud Machine Learning
Engine API
https://cloud.google.com/ml-engine/
Speaking non-
technically,
machine learning is
a process taking
input fields and
then mapping them
to a number of outputs. A classic application
is turbine data, which is used to determine
failure probabilities.
While machine-learning algorithms can be
posted on any computer able to run Python,
the amount of resources you need rises
significantly the larger the involved data
sets get. As can be inferred from the other
APIs highlighted in this feature, Google has
experience handling all kinds of machine-
learning-related jobs.
As a result, floating the job to Big G’s
servers is a convenient way to solve the
problem. In addition to removing the CPU
burden from your machines, offloading the
machine-learning payloads also improves
the learning speed – due to the large amount
of iterations it can run, the company has
been able to greatly accelerate the normally
tedious machine-learning process.
However, developers looking to get started
with machine learning must be made
painfully aware that the Cloud Machine
Learning Engine API is not a silver bullet. If
you do not understand the basics of machine
learning, your models will not work well.In a move which should surprise no one, Google lays claim to
a sizeable chunk of land in the API universe. Here are a few
especially interesting offers from Big GTHE COMPANY
HAS BEEN ABLE
TO GREATLY
ACCELERATE
THE NORMALLY
TEDIOUS
MACHINE-
LEARNING
PROCESS
FEATURES
19 groundbreaking APIs