DataSift Launches VEDO Intent Machine Learning Service

DataSiftTrack this API, a human data intelligence platform provider, has announced the launch of VEDO Intent, a new service that makes it possible for developers, marketers, and other business personnel to build advanced machine learning-based classifiers without needing a PhD or extensive data science knowledge. VEDO Intent is designed so that businesses can do more than just analyze basic social sentiment. The platform allows businesses to leverage social media data and take action based on the actual intent of their audiences. VEDO Intent includes an API (currently in early access) that developers can use to integrate platform functionality and data with third-party applications.

datasift-demo

The Keyword Relationship Model demo uses Word2Vec, D3.js, and the VEDO Intent API.

VEDO Intent uses an approach called Active Learning which is a special case of semi-supervised machine learning. VEDO Intent learns from posts as they are manually classified into categories such as rant, rave, purchase intent, or churn. VEDO Intent then builds a machine learning-based model which allows the platform to automate real-time classification of millions of social, blog, news, and other posts. This makes it possible for the platform to automatically categorize social data based on its meaning.

Developers can use the VEDO Intent API to build applications that leverage the platform's classification capabilities to discover insights hidden in social data about a brand or product. For example, an application can be built for an airline to analyze the performance of airport services such as baggage, checkin, seating, Wi-Fi, food, crew, etc. These services can be classified as rant, rave, query, feedback, lead, etc.

DataSift also announced the FORUM initiative which is a new program created to advance the application of machine learning and data-science to social media data. The program aims to enable the development of applications that utilize machine learning, deep learning, and other advanced algorithms to analyze human-created data. DataSift will publish and maintain foundational, statistical models that can be used to build and experiment with advanced analytics and algorithms.

ProgrammableWeb reached out to Tim Barker, chief product officer at DataSift who provided additional insight into VEDO Intent. He explained that the company wants to democratize machine learning so that developers and business users can easily apply the technology to human-created data. He also said that DataSift products and APIs are used almost exclusively by developers to build applications for enterprises and marketing companies.

Barker explained that VEDO Intent utilizes standard machine learning techniques which have been optimized for social data. The platform actually uses multiple algorithms including machine learning, deep learning and a popular algorithm by Google called Word2Vec. Barker said that "right now, Word2Vec is the algorithm du jour" and that "it works in an unsupervised way." The Keyword Relationship Model demo is a good example of how Word2Vec and the VEDO Intent API can be used.

Barker told ProgrammableWeb that the VEDO Intent API is currently in early access for DataSift customers and the company plans on releasing the API in a few months. The FORUM initiative will also provide APIs that will be released later this year. The APIs will provide programmatic access to the foundational, statistical models provided by DataSift.

"The industry is just in its infancy in terms of applying machine and deep learning to human-created data," said Barker. "With today's API announcement and FORUM initiative, we want to both democratize the market, and contribute back to help developers build more advanced analytical applications."

For more information about VEDO Intent and other DataSift products, visit DataSift.com

Janet Wagner is a freelance technical writer and contributor to ProgrammableWeb covering breaking news, in-depth analysis, and product reviews. She specializes in creating well-researched, in-depth content about APIs, machine learning, deep learning, computer vision, analytics, GIS/maps, and other advanced technologies.

Comments