Text Analytics API Now Available in Multiple Languages

This post is authored by Ollie Newth, Program Manager at Microsoft.

Although text often contains highly valuable data for companies, extracting meaningful data from it can be a challenge. The field of text analytics utilizes natural language processing to extract meaningful structured data from text, and often includes areas such as sentiment analysis, entity recognition and linking, and text clustering.

The Microsoft Text Analytics API  is one of the Cognitive Services that can help you turn unstructured text into meaningful insights. The API is one part of the Cortana Intelligence Suite, a family of services that helps enterprises build large-scale analytics solutions. Using a few lines of code, you can easily analyze sentiment, extract key phrases, detect topics and detect language for various kinds of text.

Check out the new demonstration experience here, also shown in the image below:

Since our launch about a year ago, we've seen customers using our service for a broad range of scenarios, all related to one theme: understanding customers' feedback in order to improve the customer experience. This data is being used to drive daily business decisions, ensuring feedback from customers remains at the core of any consumer-focused business.

We are excited to announce the launch of the following new capabilities. Sentiment analysis can now be performed in English, Spanish, French and Portuguese, and key phrases can be extracted from text in English, Spanish, German and Japanese. These features are in addition to our market-leading ability to recognize the language of text in 120 languages, and topic detection for English documents.

A common pattern which we've seen over this period has been to use text analytics alongside other offerings, such as the newly released Bot Framework and the Language Understanding Intelligent Service (LUIS). Developers may use sentiment analysis to create bots which respond appropriately to positive and negative statements. For example, user responses that have negative sentiment may indicate that a bot's response was inappropriate.

Customers such as Ziosk are using the Text Analytics API to provide a more granular understanding of their customer's experience. Al Pappa, head of Business Intelligence at Ziosk, says, "Thanks to Text Analytics by Azure Machine Learning, we are able to incorporate guest sentiment into our actionable guest feedback platform that delivers a comprehensive view of guest satisfaction and server performance."

All the capabilities can be tested using our demo experience, and you can get started for free by signing up in the Azure Portal.

We are always looking for feedback, so please reach out in the comments below or submit ideas using the Cognitive Services UserVoice.

Ollie