http://www.digitaljournal.com/tech-and-science/technology/google-adds-content-classification-features-to-its-language-cloud/article/502843

Google adds content classification features to its language cloud

Posted Sep 19, 2017 by James Walker
Google has announced the launch of content classification support on its Natural Language Cloud. The addition will particularly benefit large brands and media companies wanting to work out how consumers are talking about their products.
The update will give companies even deeper insights into their data
The update will give companies even deeper insights into their data
Pexels / Negative Space
Google turned on the new features today. Its Cloud Natural Language API provides AI-powered services to help developers extract information from written material. The cloud can organise content, deliver insights and now automatically sort it.
Content classification
The content classification features give Cloud Natural Language the ability to categorise documents into over 700 distinct labels. These range from broad umbrella terms, such as "Arts & Entertainment" and "Law & Government," to narrower niche phrases more applicable to individual industries.
For companies using Google's Cloud, this addition is particularly notable. They can now pass articles about their brand or products through Cloud Natural Language to see how people are referencing them. The cloud infers the overall theme of the article, allowing the company to understand which industries are benefitting the most from its products.
Entity sentiment analysis
Google's also announced improvements for its existing sentiment analysis capabilities. Sentiment analysis is one of Cloud Natural Language's most popular features. It uses AI to interpret the overall sentiment expressed in a text. Brands could use it to work out whether new products are being reviewed favourably or assess how their social media campaign is faring.
With today's update, sentiment analysis now offers more precise controls that extend down to individual entities. Instead of restricting analysis to whole sentences or paragraphs, the cloud can now parse the sentiment of specific places and things mentioned in the text. Companies can start to look at the sentiment expressed to their specific product, rather than the overall sentiment of the larger article it's mentioned in.
Unlocking data
Combined, these two additions give brands a lot more power to unlock insights within written information. Cloud Natural Language can filter through different sentiments and then categorise it by subject.
READ NEXT: APIs recurring theme in digital transformation
For companies dealing with multiple industries, this offers more immediately actionable data to help finetune their product. If a software firm sees they're getting positive reviews from enterprises but a cooler response from education, they can adapt their roadmap to plan new education features.
Google said entity sentiment analysis is available today to all Cloud Natural Language users. Content classification is initially being offered in beta but will expand broadly in the future. Google said the features will help businesses "get the most from their data," helping them digest the hidden insights in their data pools.