Text Analytics API (that is Microsoft Azure’s comprehensive text analytics approach) is a new Microsoft cognitive service wherein a few lines of code can help in sentiment analysis, key phrase extraction, topic detection and language detection. The API employs advanced natural language processing techniques to deliver predictions that are important for the businesses of today.
The need of this API comes from the growing competition where every business is chasing customers, trying to attract news ones, trying to retain existing ones and trying to market something more than just their product/ service. Traditional marketing ways are not working to help any of the above. To expand market share, customer intelligence is important and, in the recent times, it has become a norm rather than just a requirement. The sources where important customer information lies has changed over the last few years. Traditional places where customer data was available is long gone. Social media, call logs, blog comments, product reviews, online interaction are some of the most important sources which have become the gold mine of customer data and once analyzed it becomes customer intelligence.
If done manually, the process is time-consuming and can be full of human errors. By the time the manual process is over, there are other important trends that are changing and key discussions that are getting missed. Higher chances of human error can (most likely) jeopardize a huge business opportunity. With text analytics (which is a machine learning activity using big data), extracting value out of tons of virtual data becomes a matter of a few minutes. Smart enterprises are able to leverage the data and create an actionable feedback platform from the results derived from the new cognitive service of Microsoft.
The new API enables can turn a mass of words into structured content. For example, if there is an organized way wherein your customer feedback, social media and blog comments, product reviews etc. is presented, it is easier to analyze and draw inferences from that data. The analysis can help you improve an existing marketing strategy, understand what is not liked about your product, consider how your brand is perceived, and plan a future marketing strategy. It is a detailed, managed-services solution that has all the right mechanisms (including information management, storage, intelligence services, machine learning and reporting).
Microsoft has provide a link wherein you can get a demo of the Text Analytics API. You can even check an example of sentiment analysis, key phrase extraction and language detection. The first 5,000 transactions are free and afterwards different plans are available along with different pricing options. Guide for developers is available as documentation and API reference.
As per a guide on API, this is a newly released interface, which returns the top detected topics for a list of submitted text records.
Authorization: Basic <creds>
Where <creds> = ConvertToBase64(“AccountKey:” + yourActualAccountKey);
Single Response APIs
Topic Detection APIs
Topics – Submit job
Topics – Poll for job results
By: Meenakshi Vashisht