Interests-based Bidirectional Fusion of Web Knowledge Retrieval and Reasoning

Supported by:


Project Overview

In the cyber-physical world, the World Wide Web and its various types of users collectively form an ¡°open complex giant system [1]¡±. The user base of the Web platform is growing very fast and the Web knowledge sources have been going to infinite scale, which leads to the fact that each user¡¯s individual difference need to be realized so that the Web knowledge processing systems can be more user driven and user oriented.

This project aims at investigating into the problem of diversity of user requirements and scalability for Web knowledge processing through user interests-based bidirectional fusion of knowledge retrieval and reasoning. The main efforts of this project will go into the following direction:

  1. We focus on distributed user interests resources integration, fusion algorithms as well as qualitative and quantitative interests analysis methods on the fly and from the Web.
  2. With exploring the structural and dynamical characteristics of interests as the foundation, we investigate into the semantic representation of interests, and transforming the interests analysis results into semantic knowledge sources.
  3. We propose and investigate into the interest logic from the semantic, temporal, spatial, and structural perspectives of human interests under the social environment, which serves as the foundation for developing retrieval and reasoning fusion strategies.
  4. Based on the analysis results for user interests and the interest logic, we investigate into the concrete strategies for retrieval and reasoning bidirectional fusion (improving retrieval by reasoning, and improving reasoning by retrieval), with user evaluations based on real-world applications.

The developed platform based on the theoretic efforts in this project will support the development of Web-enabled personalized semantic search and reasoning engines, which serve as infrastructures to turn ¡°The (semantic) Web¡± to ¡°Your (semantic) Web¡±.



[Usecase 1]
The large scale scientific literature semantic search and analysis system as well as the active academic collaboration and visit recommendation system to be developed under this project will provide many useful supporting functionalities to assist researchers in conducting scientific research. In addition, the developed system is aimed at supporting the research on the Science of Team Science [2].

Now the very preliminary test version of the sub system "Active Academic Visit Recommendation Application (AAVRA)" is in prototype shape. It is based on the LarKC platform, with 3 workflows (including 7 plugins running within these workflows).

A Screenshot of the "Active Academic Visit Recommendation Application (AAVRA)"

[Usecase 2] To be added¡­


Related History of This Project

The idea of "unifying reasoning and search" was firstly proposed in [3], and initial research on selecting user interests related semantic data to reduce the problem space for reasoning was done as an effort for and contribution to the Large Knowledge Collider Project (EU FP-7 215535). The idea and agenda for ¡°Interests-based Bidirectional Fusion of Knowledge Retrieval and Reasoning¡± is a reconsideration and plan for combining the end users and the Web knowledge processing platform into an interactive ¡°meta-synthesis system [4]¡±.
Important Publications of previous work within the LarKC project are selectively listed:


[1] Xuesen Qian, Jingyuan Yu, Ruwei Dai. A New Discipline of Science: Open Complex Giant System and Its Methodology. Chinese Journal of Systems Engineering & Electronics, 1993, 4(2):2-12.
[2] Fei-Yue Wang. From AI to SciTS: Team Science and Research Intelligence. IEEE Intelligent Systems 26(3): 2-4 (2011)
[3] Dieter Fensel, Frank van Harmelen. Unifying Reasoning and Search to Web Scale. IEEE Internet Computing 11(2): 94-96 (2007)
[4] Ruwei Dai. Meta-synthesis from Qualitative to Quantitative Approach. Chinese Journal of Pattern Recognition and Artificial Intelligence, 1991, 4(1): 5-10.

Facts about This Project

Project Leader: Yi Zeng
Funded By: the National Natural Science Foundation of China (NSFC)
Grant Type: the Young Scientists Fund
Grant Number: 61100128
Duration: 36 months


  • Yi Zeng, Hongwei Hao, Ning Zhong, Xu Ren, Yan Wang. Ranking and Combining Social Network Data for Web Personalization. Proceedings of the International Workshop on Data-driven User Behavioral Modelling and Mining from Social Media, co-located with the 21st ACM International Conference on Information and Knowledge Management (CIKM 2012), Hawaii, USA, October 29th, 2012.
  • Yi Zeng, Danica Damljanovic, Xu Ren, Yan Wang. Identification and Selection (In Chinese). In: Zhisheng Huang and Ning Zhong (Eds.), Scalable Semantic Data Processing: Platform, Technology and Application, Higher Education Press, 2012.