The ACM SIGKDD Conference on Knowledge Discovery and Data Mining (ACM SIGKDD) is the top international conference in the field of data science. It is coordinated by the Association of Computing Machinery (ACM). KDD brings together multi-disciplinary data science, data mining, knowledge discovery, and big data experts and practitioners from around the world each year. In 2016, more than 3,000 people participated. As a top event in the field of data science, the papers and speeches published on KDD represent the latest developments in the field of data science and the most cutting-edge applications, with strict screening criteria. In 2016, KDD attracted a total of 1115 paper submissions, and the overall recruitment rate was less than 20%. All the presentations were directed invitations. 2016 ACM SIGKDD Data Mining and Knowledge Discovery Conference (abbreviated as KDD Conference) kicked off in San Francisco. As the top event in the field of data science, 2016 KDD Conference brought together Stanford, University of California at Berkeley, Carnegie Mellon University, and Tsinghua University. Leading research scholars such as top universities and representatives of leading companies in the global data science field discuss the latest research findings, commercial application innovations and trends in data science.
Manager Luo Zheng invited to give a speech at the conference
On August 16th, Luo Zheng, general manager of Tencent’s social advertising department, was invited to deliver a speech at the conference — “Building User Profiles from Online Social Behaviors, with Applications in Tencent Social Ads.†"Applications in social advertising"), in-depth interpretation of how Tencent social advertising through data analysis more accurate and comprehensive understanding of users, enhance user advertising experience, for the brand and the user to establish a more valuable connection.
Precipitation and deep plowing of social behavior data
Luo Zheng pointed out that Tencent aims to “enrich people’s Internet life with leading technology†and continuously drives innovation in the five major areas of social networking, entertainment, information, tools, and payment, and enriches the user’s online life experience. Tencent’s products affect the daily lives of Chinese netizens. Every bit, lay a solid foundation for the accumulation of Tencent social behavior data. Especially in terms of basic attribute data, social relationship chain data, and LBS data, Tencent is far ahead in terms of quantity and quality.
At the meeting, Luo Zheng detailed how Tencent tapped social behavior data to accurately construct user portraits. Tencent has built a leading technology tool in the process of mining social behavior data. Peacock's large-scale hidden topic learning tool can efficiently decompose 500 million x 100 million ultra-large-scale matrices, providing powerful features for precise mining, and in user interests. Deep learning models are used directly in mining and specific population identification to improve accuracy. In terms of how to lay a solid foundation in the data and further deepen user understanding, Luo Zheng believes that Tencent's strong account system can effectively aggregate multi-screen data, break data islands, optimize data quality, and expand user attribute coverage.
From user understanding to data marketing applications
Luo Zheng said that advertisers can use Tencent's precise user insight capabilities to target users through geographic location, device behavior, and vertical industry-related features and behaviors. Tencent's user portrait building capabilities can help brands reach more high quality The crowd can also provide users with a better advertising experience, which in turn improves user conversion rates.
Based on the precise targeting of existing target users, advertisers can also build user demographics based on seed demographics through the “Lookalike†technology of Tencent’s social advertising, and discover more similarity and highness among seed users. Related people.
At the same time, Tencent Social Advertising has built an industry-leading data management platform - Data Management Platform (DMP). By opening Tencent's big data capabilities, it will gather together scattered data across the industry, create connectors for big data, integrate vertical industry data, and work together with partners like JD.com to build a quality data ecosystem.
Luo Zheng’s speech brought unique insights to data scientists and practitioners on data mining and application of social behavior. Luo Zheng stated that Tencent’s social advertising will continue to focus on deeper data accumulation and continuous improvement of data technology to drive marketing efficiency. The mission is to bring the best advertising experience to the user and help the brand maximize the value of product value.
Via: original report