BAT: China’s Three Big Data Leaders

We examine the “three big mountains” in Chinese Internet and Big Data industry: Baidu, Alibaba, and Tencent (together called BAT), and look into their different strategy and focus.

By Liyang Tang, Aug 2014.

In Chinese Internet industry, there are three big leaders in the big data wave: Baidu, Alibaba, and Tencent, called BAT. From the data source perspective, Baidu collects the data based on user search, Alibaba owns the transaction data and credit data, and Tencent has the social network data.

BAT: China's Three Big Data Leaders
As a search engine provider, Baidu is collecting traffic data. Baidu has a strong technical support, and tries to present a new picture of intelligent search for users and developers through big data, cloud computing, deep learning and other core technologies. Specifically, the strategy of Baidu includes: open cloud, data factory and Baidu brain. Based on an open big data platform, a big data factory is established. After that, deep learning algorithms, data models and massive GPU parallel computing techniques are integrated through Baidu brain. Technically, there are three laboratories in Baidu: Silicon Valley Artificial Intelligence Laboratory, Deep Learning Institute (IDL) and Beijing Big Data Lab.

Alibaba was first developed from B2B services of SMEs, and then extended to B2C by Taobao and Alipay. As a big e-commerce enterprise, Alibaba owns the most valuable “golden” data. Therefore, the most important thing for Alibaba is the mastering of data. To achieve this, the following measures have been taken:
  1. ensure data security (e.g. protection of privacy and individual businesses);
  2. ensure the quality of data (e.g. remove false data);
  3. provide uniform data with the same standards across various departments (such as conversion rate);
  4. refine the original data (e.g. in line with the business application scenarios);
  5. access to external data (such as cooperation with other platforms);
  6. establish a data committee. However, Alibaba is more working on building the underlying architecture for data collection and sharing, than performing deep analyzing and data mining.

Tencent has the biggest social data of China. Tencent is good at producing products. Currently, Tencent is trying to integrate the backend data from Qzone, Wechat and e-commerce products, and establishes a stable ecosystem centered by users. However, Tencent has not yet focused on big data techniques. Besides, there is a lack of big data technology leaders in Tencent.

In summary, Baidu has put more efforts in exploring big data techniques, including deep learning, big data analyze, Human-Computer Interaction (HCI), 3D vision, heterogeneous computing and image recognition, etc; Alibaba builds a mature system of collecting, processing, storing and managing data, but still lacks of data mining attempts; Tencent has a stable product ecosystem, but remains a conservative manner in big data analysis.

Liyang Tang, @tangliyang, is studying for a Ph.D in data mining and business intelligence in China.