Have you ever wondered if there’s a way to quantify how customers collectively feel about your products based on online reviews? Alternatively, perhaps you’ve wondered if there’s a way to systematically protect all your content from profane language and hate speech? With the help of NLP Analytics, you can accomplish all the above, uncovering important customer insights, more easily moderating your online content, and much more. In this article, we’ll take a closer look at three of our five NLP Analytics APIs and how they can make a big difference for your business.

Sentiment Analysis & Classification API

The purpose of Sentiment Analysis is to find out how a certain person – or group of people – feels about a given subject based on how they write about it. Are their feelings positive, negative, or neutral? With a Sentiment Analysis API in your toolbelt, you can gain a better understanding of how your customers feel about your products and make better informed marketing decisions as a result. Its use isn’t limited to only customer centric businesses, however – there are relevant applications of this technology in just about any professional sector you can think of.

To quantify written feelings, this advanced API rigorously analyzes language in a body of text and provides a Sentiment Classification Score value between -1.0 and 1.0. Scores closer to -1.0 indicate a negative view on a given subject, while values closer to 1.0 indicate a positive view; scores closest to zero instead suggest a more neutral sentiment. This API additionally returns a Sentiment Classification Result string based on the score provided; the possible responses are positive, negative, or neutral.

Profanity & Obscenity Detection API

Bad language is a bad representation of your brand – even if it comes from voices outside your organization. When customers come away feeling disappointed, aggrieved, or even elated from an experience with any of your products, they can occasionally express themselves with inappropriate language, which – especially in the case of online reviews – can create an unsavory environment for new patrons. This API performs a practical content moderation service by employing advanced analysis to detect profane/obscene language from text; it returns a Profanity Score Result between 0.0 – 1.0, with values closer to 0.0 suggesting very little profanity, and values closer to 1.0 indicating a high degree of profanity.

Hate Speech Detection API

It goes without saying that content containing Hate Speech should have no place anywhere on any company’s platform. Considering the overwhelming importance of flagging and removing Hate Speech content – paired with the difficulty human content moderation teams face in scouring insurmountable written materials online – Hate Speech Detection can make a huge difference, reducing the burden of manual review through automatic detection of discriminatory language in various forms. Much like the previously described Profanity Detection API, this API provides a Hate Speech Score Result between 0.0 and 1.0, with values on the lower end of the spectrum indicating very little (or no) hate speech was detected, and scores on the higher end of the spectrum suggesting Hate Speech was, in fact, present. Values above 0.8 indicate a particularly high likelihood that Hate Speech was rampant in a body of text, making it easy for content moderation groups to efficiently triage their response.