Files in this item

FilesDescriptionFormat

application/pdf

application/pdfKin Hou_Lei.pdf (4MB)
(no description provided)PDF

Description

Title:ONet: Search, explore, and visualize hierarchical topic summarization on Twitter using network-OLAP
Author(s):Lei, Kin Hou
Advisor(s):Han, Jiawei
Department / Program:Computer Science
Discipline:Computer Science
Degree Granting Institution:University of Illinois at Urbana-Champaign
Degree:M.S.
Genre:Thesis
Subject(s):Heterogeneous Information Network
Online Analytical Processing
Abstract:Multi-dimensional data is one of the most abundant and publicly available data sources for people to access today. It forms hierarchy naturally(e.g., people publish posts about different topics or share information from various countries. Given a topic such as technology, it many form sub-topics like mobile devices and web news. Similarly, given a country, it naturally forms hierarchy such as states, counties, cities, and towns). Multi-dimensional data may also interlink together with URLs, images, videos and people to form a rich heterogeneous information network. In this study, we propose ONet to search, explore and visualize hierarchical summarization on multi-dimensional data. In particular, we take Twitter data as an example to show the power of integration of data warehouse and OLAP technologies with information network. Based on Twitter data, ONet summarized important events at different granularities. With the interlinked events and other entities consisted in the network, we investigated some state-of-the-art ranking algorithms, and developed a ranking model in a learning-to-rank approach to rank heterogeneous entities. Experimental results on a large scale real data set show that our proposed ranking model achieves high efficiency and outperforms all compared baselines.
Issue Date:2014-05-30
URI:http://hdl.handle.net/2142/49348
Rights Information:Copyright 2014 Kin Hou Lei
Date Available in IDEALS:2014-05-30
Date Deposited:2014-05


This item appears in the following Collection(s)

Item Statistics