| about | research | publications | cv |
YU-RU LINIQSS, Harvard University & CCS, Northeastern University Office: Phone: 617-373-4222 Email: yuruliny@gmail.com |
 |
about
I am a postdoctoral research fellow at Institute for Quantitative Social Science, Harvard University and College of Computer and Information Science, Northeastern University. My research interests have been in analysis and visualization of interpersonal activities in social networks - in particular, large-scale community dynamics, high-dimensional social information summarization and representation. Currently I am working with Dr. David Lazer on understanding the structural and dynamical aspects of human communities related to political choices and under unfamiliar situations.
I received my Ph.D. in Computer Science with a concentration in Arts Media and Engineering from Arizona State University, where I worked with my advisor Dr. Hari Sundaram and Dr. Aisling Kelliher. My Ph.D. work focused on extracting human communities that collaborate around certain topics or shared media artifacts. I have proposed matrix and tensor based techniques for analyzing community structures and evolutions in time-varying heterogeneous social networks, and developed visualizations to support community discovery in the context of everyday social media use.
updates
04/2012 I will be giving a talk at Arts, Humanities, and Complex Networks - 3rd Leonardo satellite symposium at NetSci2012.04/2012 Our paper "Recipe recommendation using ingredient networks" was accepted in Web Science 2012. It was featured in Science News.03/2011 Our paper "More Voices than Ever? Quantifying Bias in Social and Mainstream Media" has been accepted to ICWSM 2011. This paper presents a systematic, empirical comparison between social (blog) and mainstream (news) media, on the extent and dynamics of media biases. Our study reveals distinct characteristics in the two media.11/2011 We have posted a series of election related visualizations, including campaign wordcloud, issue meter, contribution cartogram, and a live analysis of political topics on Twitter. Read more about this project inpress.
research
My publications and talk slides can be downloaded here.» Current projects «
bias in social and mainstream media
Social media, such as blogs, have played an increasingly important role in shaping our society and world-wide politics. They are now seen as democratic entities that allow more voices to be heard than the conventional mainstream media as well as a balancing force against the arguably slanted elite media, which redefine the information available to the public and can affect public opinion and decision-making. I believe one way to understand the impact of social media is to systematically compare them with their mainstream media counterparts. In our study we propose empirical measures to quantify the extent and dynamics of "bias" in social (blog) and mainstream (news) media. The measurements are not normative judgments, but examine bias by looking at the attributes of those being mentioned, against a null model of "unbiased" coverage. Our current findings suggest distinct characteristics in news and blog media, and reveal differences in the process of coverage selection between the two media. See related publication.mesoscopic structure and social aspects of human mobility
The individual movements of large numbers of people are important in many contexts, from urban planning to disease spreading. Datasets that capture human mobility are now available and many interesting features have been discovered, including the ultra-slow spatial growth of individual mobility. However, the detailed substructures and spatiotemporal flows of mobility--the sets and sequences of visited locations--have not been well studied. We show that individual mobility is dominated by small groups of frequently visited, dynamically close locations, forming primary "habitats" capturing typical daily activity, along with subsidiary habitats representing additional travel. These habitats do not correspond to typical contexts such as home or work. The temporal evolution of mobility within habitats, which constitutes most motion, is universal across habitats and exhibits scaling patterns both distinct from all previous observations and unpredicted by current models. The delay to enter subsidiary habitats is a primary factor in the spatiotemporal growth of human travel. Interestingly, habitats correlate with non-mobility dynamics such as communication activity, implying that habitats may influence processes such as information spreading and revealing new connections between human mobility and social networks. Read more.the geography of money and politics
How is money in politics mobilized? Much of the literature on political behavior focuses on individual socioeconomic attributes (e.g., income, education, ethnicity, etc.) as predictors of behavior. In this project we examine the social antecedents for contributing to campaigns, with a particular focus on the role of population density in facilitating contributions to campaigns. We hypothesize that the idea of contributing to a campaign is easier to spread in a densely populated region, where the daily opportunity of individuals being exposed to the same idea via their social network is high, compared with people living in a less populous region. Furthermore, the effect of population density is inhomogeneous with respect to the dynamics of residents' mobility. We test our hypothesis on a 10-year US campaign contribution data extracted from the records provided by the Federal Election Commission (FEC), with additional socioeconomic information and commuting flow data from the 2000 Census. Our results suggest that, controlling for the wealth-related variable of a given region, population density and outgoing commuting flow account for a significant amount of the variation in contribution levels. This analysis partially addresses a perennial puzzle in US politics: how Democrats remain competitive in the money race in US politics, despite that wealth and income tend to be associated with being Republican. Our analysis implies that more densely populated areas tend to be Democratic, and thus the Democratic party starts with a substantial advantage in mobilizing monetary support.critical social ties from change of situations
The recent available large-scale behavioral data such as cell phone call records offers unprecedented opportunities to study social relationships. The effect of social relationships (or ties) on innovation, knowledge sharing, the spread of opinions and behaviors has become increasingly clear. What is less clear is how to identify critical relationships among the many unimportant relationships from the cacophony of information from various sources. We propose a general approach to reveal critical ties through a natural experiment, where an exogenous event such as emergencies (earthquakes, bombings, etc) occurs should lead individuals to communicate with certain types of ties. The experiment allows us to examine questions including: what are the behavioral correlates of emergency calls? how does this vary with demographic factors, and how does it deviate from the "average" behavior of individuals? By building predictive models on country-wide mobile phone records, we discover key factors related to people's calling decisions. Our findings suggest people's effective social networks change with different situation, and are affected by context-dependent factors such as geographical distance. An important implication of our study is that it is now possible to predict a large-scale network dynamics, such as information propagation, by transforming the set of micro-level network elements (including individuals' present multiplex ties interacting with context factors) at similar conditions.
» Previous projects «
- community discovery in dynamic, rich media social networks How can we extract communities from data of interactions with rich contexts and track the dynamics of membership and topics of interests within communities? We propose multi-tensor factorization framework for analyzing the dynamics of heterogeneous social networks.
- summarization of time-evolving, social media stream How can we characterize time-evolving patterns in community-shared media? We propose a joint matrix factorization framework that incorporates image content features and contextual information to discover distinct temporal patterns of group photo streams.
- adversarial information retrieval How can we combat spam in the blogosphere? Our spam blog (splog) methods combine traditional content based features with temporal and link signatures with excellent results.