Personalized search over the streaming and heterogeneous social media data

The primary goal of this project is to advance a new class of search engines to provide social network users with the ability to harness the growth of data, where both internal user-generated data and external web data can be systematically captured, continuously retrieved and summarized.

Project title: Personalized search over the streaming and heterogeneous social media data

Project dates: ​2018 onward

Grants and funding: ​Google Faculty Research Award

Description

This project is done by the Big Data & Database Group at RMIT led by A/Prof. Zhifeng Bao.

Social networks like Facebook and Twitter are hubs for users to acquire up-to-date information on the Internet. Apart from information generated within social networks, external information (from the Internet) of higher quality and selectivity is shared by users in social networks and can be accessed as well. The primary goal of this project is to advance a new class of search engines to provide social network users with the ability to harness the growth of data, where both internal user-generated data and external web  data  can  be  systematically  captured,  continuously  retrieved  and  summarized.  Organizing Internet content around social media will create a novel way to search for contents on the Internet compared to conventional search engines.

Research strategy

It designs lightweight and update-efficient indexes to manage heterogenous data that is evolving temporally. 

It investigates novel data structures to capture the lifespan and propagation of microblog entries, thereby capturing a story line with more content and relationships of related microblogs, which are missed in existing search engines that only find an independent list of individual microblogs.  

It develops novel summarization techniques to preserve important properties  such  as  the distinguishable textual contents, propagation patterns in Social Network Platforms and social popularity of the story line.

Rationale

This project uses the connectivity of the internet as a means to develop dynamic congregation and easy access to all forms of user-generated data, including the content generated in social media and the content of conventional websites shared via social media.

Key people

  • Associate Professor Zhifeng Bao

Associated journal publications

  • Shixun Huang, Zhifeng Bao, Guoliang Li, Yanghao Zhou, J. Shane Culpepper: Temporal Network Representation Learning via Historical Neighborhoods Aggregation. IEEE ICDE 2020: 1177-1128. [CORE A*]
  • Shixun Huang, Zhifeng Bao, J. Shane Culpepper, Bang Zhang: Finding Temporal Influential Users Over Evolving Social Networks. ICDE 2019: 398-409. [CORE A*]
  • Yipeng Zhang, Yuchen Li, Zhifeng Bao, Songsong Mo, Ping Zhang: Optimizing Impression Counts for Outdoor Advertising. KDD 2019: 1205-1215. (Best Paper Award Runnerup) [CORE A*]
  • Ping Zhang, Zhifeng Bao, Yuchen Li, Guoliang Li, Yipeng Zhang, Zhiyong Peng: Trajectory-driven Influential Billboard Placement. KDD 2018: 2748-2757. [CORE A*]
  • Ji Sun, Zeyuan Shang, Guoliang Li, Zhifeng Bao, Dong Deng: Balance-Aware Distributed String Similarity-Based Query Processing System. PVLDB 12(9): 961-974 (2019). [CORE A*]
aboriginal flag
torres strait flag

Acknowledgement of country

RMIT University acknowledges the people of the Woi wurrung and Boon wurrung language groups of the eastern Kulin Nation on whose unceded lands we conduct the business of the University. RMIT University respectfully acknowledges their Ancestors and Elders, past and present. RMIT also acknowledges the Traditional Custodians and their Ancestors of the lands and waters across Australia where we conduct our business. - Artwork created by Louisa Bloomer