Alibaba unveils big data cloud platform

Chinese internet giant's cloud business, Aliyun, launches what it's touting to be "China's first artificial intelligence platform", which will facilitate analysis of large volumes of data to predict user behavior.
Written by Eileen Yu, Senior Contributing Editor

Alibaba Group's cloud computing business, Aliyun, has unveiled what it is touting to be "China's first artificial intelligence platform" that will enable developers to analyse large volumes of data.

Dubbed DT PAI, the new offering features drag-and-drop functions and will generate forecasts of user behaviour and industry trends, Aliyun said on Tuesday. It runs on Alibaba's core algorithm structure and includes feature engineering, large-scale machine learning, as well as deep learning. The Chinese vendor said Alibaba's Open Data Processing Service (ODPS) is capable of handling 100PB data, the equivalent of 100 million HD films, in six hours.

According to Aliyun, DT PAI can be operated by linking standardised modules and setting the targeted parameters prior to application development, bypassing the need for traditional coding.

"Our goal is to create a one-stop artificial intelligence development, publishing, and sharing platform through data, calculations, and data connections," said Wei Xiao, Aliyun's senior product manager. "In the past, the field of artificial intelligence was only open to a very small number of qualified developers and required the use of specialised tools. Such an approach was prone to error and redundancy."

With DT PAI, Wei said developers with little or no experience in this area would be able to construct a data application from scratch in a shorter period of time. "What used to take days can be completed within minutes," he added.

Aliyun said insights generated from the platform could be applied to daily life, pointing to how shoppers would be able to take a photo of an outfit and automatically be redirected to the product's profile page on Alibaba's e-commerce website, Taobao Marketplace.

Editorial standards