Saturday, June 24, 2023

Serendipitous Convergence of the Dodgers and Tarzan Boy

On a few occasions this year, I had noticed the Dodger organist playing Tarzan Boy, and I was wondering why.

I had thought it was for something like certain leaping catches in the outfield, but it turns out that it is specifically for rookie James Outman. I figured that out last night because they played it when he got a hit in his first at bat. (I revised my opinion from seeing this when Outman was at the plate and running for a hit, with his locks flowing.) I thought it might have been because of his luxuriantly flowing hair.

I decided to google it to confirm whether I was right, and indeed Tarzan Boy is played specifically for good James Outman action, although it seems to actually be because of a nickname that is catching on. (I hadn't known about eh nickname.)

I am very amused by the fact that this is a convergence between the Dodgers and Tarzan Boy, given how many people from my Lloyd House days at Caltech would associate each of those two things individually with me.

What Happens in Dallas Stays in Dallas

Well, unfortunately, I won't be making my connection (annoying flight delay), and I will be staying an unintended night in Dallas before resuming my journey in the morning.

But at least I won't be liveblogging from the Dallas airport, as I did 16 years ago.

Wednesday, June 21, 2023

"Bounded-Confidence Model of Opinion Dynamics with Heterogeneous Node-Activity Levels"

One of my papers came out in final form today. Here are some details.

Title: Bounded-Confidence Model of Opinion Dynamics with Heterogeneous Node-Activity Levels

Authors: Grace J. Li and Mason A. Porter

Abstract: Agent-based models of opinion dynamics allow one to examine the spread of opinions between entities and to study phenomena such as consensus, polarization, and fragmentation. By studying models of opinion dynamics on social networks, one can explore the effects of network structure on these phenomena. In social networks, some individuals share their ideas and opinions more frequently than others. These disparities can arise from heterogeneous sociabilities, heterogeneous activity levels, different prevalences to share opinions when engaging in a social-media platform, or something else. To examine the impact of such heterogeneities on opinion dynamics, we generalize the Deffuant-Weisbuch (DW) bounded-confidence model (BCM) of opinion dynamics by incorporating node weights. The node weights allow us to model agents with different probabilities of interacting. Using numerical simulations, we systematically investigate (using a variety of network structures and node-weight distributions) the effects of node weights, which we assign uniformly at random to the nodes. We demonstrate that introducing heterogeneous node weights results in longer convergence times and more opinion fragmentation than in a baseline DW model. The node weights in our BCM allow one to consider a variety of sociological scenarios in which agents have heterogeneous probabilities of interacting with other agents.

"Lonely Individuals Process the World in Idiosyncratic Ways"

One of my papers that came out a couple of months ago now also has its final volume and page numbers. Here are some details about the article.

Title: Lonely Individuals Process the World in Idiosyncratic Ways

Authors: Elisa C. Baek, Ryan Hyon, Karina López, Meng Du, Mason A. Porter, and Carolyn Parkinson

Abstract: Loneliness is detrimental to well-being and is often accompanied by self-reported feelings of not being understood by other people. What contributes to such feelings in lonely people? We used functional MRI of 66 first-year university students to unobtrusively measure the relative alignment of people’s mental processing of naturalistic stimuli and tested whether lonely people actually process the world in idiosyncratic ways. We found evidence for such idiosyncrasy: Lonely individuals’ neural responses were dissimilar to those of their peers, particularly in regions of the default-mode network in which similar responses have been associated with shared perspectives and subjective understanding. These relationships persisted when we controlled for demographic similarities, objective social isolation, and individuals’ friendships with each other. Our findings raise the possibility that being surrounded by people who see the world differently from oneself, even if one is friends with them, may be a risk factor for loneliness.

Sunday, June 18, 2023

Thursday, June 08, 2023

"Detecting Political Biases of Named Entities and Hashtags on Twitter"

One of my papers came out in final form earlier today. Here are some details. (This is in collaboration with computer scientists, and stylistically it is rather different from much of my work. However, you'll still notice my hand in it. :P)

Title: Detecting Political Biases of Named Entities and Hashtags on Twitter

Authors: Zhiping Xiao, Jeffrey Zhu, Yining Wang, Pei Zhou, Wen Hong Lam, Mason A. Porter, and Yizhou Sun

Abstract: Ideological divisions in the United States have become increasingly prominent in daily communication. Accordingly, there has been much research on political polarization, including many recent efforts that take a computational perspective. By detecting political biases in a text document, one can attempt to discern and describe its polarity. Intuitively, the named entities (i.e., the nouns and the phrases that act as nouns) and hashtags in text often carry information about political views. For example, people who use the term “pro-choice” are likely to be liberal and people who use the term “pro-life” are likely to be conservative. In this paper, we seek to reveal political polarities in social-media text data and to quantify these polarities by explicitly assigning a polarity score to entities and hashtags. Although this idea is straightforward, it is difficult to perform such inference in a trustworthy quantitative way. Key challenges include the small number of known labels, the continuous spectrum of political views, and the preservation of both a polarity score and a polarity-neutral semantic meaning in an embedding vector of words. To attempt to overcome these challenges, we propose the Polarity-aware Embedding Multi-task learning (PEM) model. This model consists of (1) a self-supervised context-preservation task, (2) an attention-based tweet-level polarity-inference task, and (3) an adversarial learning task that promotes independence between an embedding’s polarity component and its semantic component. Our experimental results demonstrate that our PEM model can successfully learn polarity-aware embeddings that perform well at tweet-level and account-level classification tasks. We examine a variety of applications—including a study of spatial and temporal distributions of polarities and a comparison between tweets from Twitter and posts from Parler—and we thereby demonstrate the effectiveness of our PEM model. We also discuss important limitations of our work and encourage caution when applying the PEM model to real-world scenarios.