Information Network Modeling and Mining
thesisposted on 27.10.2017 by Jingyuan Zhang
In order to distinguish essays and pre-prints from academic theses, we have a separate category. These are often much longer text based documents than a paper.
We are living in an Internet world with huge amounts of information. For example, online Question-and-Answer (Q&A) websites include different kinds of informational objects, i.e., questions, answers and users. Web search logs contain queries and clicked webpages. These online infrastructures usually represent information in various forms. An urgent challenge is to discover meaningful knowledge from massive information that can empower online platforms from different perspectives. Motivated by this trend, my research addresses the modeling and mining tasks for several online platforms, including the Q&A websites, the search engines and the social media. Proposing information network based modeling and mining framework regarding to heterogeneous data sources is the major goal of my research. The frameworks, by extracting interconnected objects, modeling them into information networks, and designing network based learning algorithms, help improve the service offered by different online platforms. The methodology of the information network based modeling and mining is proved to be effective on a series of knowledge discovery topics, including the co-ranking problem in large-scale Q&A sites, the learning of entity types from massive search query logs, and the detection of emerging relationships from news and knowledge graphs. The future work is to explore the network base modeling and mining techniques on more online platforms and study how they can fit for new situations.