Hope vs. Hate: Understanding User Interactions with LGBTQ+ News Content in Mainstream US News Media through the Lens of Hope Speech

Jonathan Pofcher, Christopher M. Homan, Randall L. Sell, Ashiqur R. KhudaBukhsh · 2025

Read Hope vs. Hate: Understanding User Interactions with LGBTQ+ News Content in Mainstream US News Media through the Lens of Hope Speech by Jonathan Pofcher, Christopher M. Homan, Randall L. Sell, Ashiqur R. KhudaBukhsh (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-09004.txt.

Introduction

Jonathan Pofcher and co-authors study how users react to LGBTQ+ news content on major United States news channels, in a paper from 2025. Their lens is hope speech: comments that support or uplift, set against hate and neutral talk.

The team gathered well over a million comments on thousands of YouTube news videos. They built a fine-grained classifier for the four comment types. A balanced annotation study, guided by a public health expert on LGBTQ+ health, produced a labeled dataset with rater details kept. Ratings turned out to track the rater's political beliefs strongly. Models trained on one group's labels disagreed with others in the wild. Large models used cold aligned more with liberal raters.

Catalog record

CREATORJonathan Pofcher, Christopher M. Homan, Randall L. Sell, Ashiqur R. KhudaBukhsh
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE81 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-09004

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Hope vs. Hate: Understanding User Interactions with LGBTQ+ News Content in Mainstream US News Media through the Lens of Hope Speech Jonathan Pofcher, Christopher M. Homan, Randall L. Sell, Ashiqur R. KhudaBukhsh Abstract This paper makes three contributions. First, via a substantial corpus of 1,419,047 comments posted on 3,161 YouTube news videos of major US cable news outlets, we analyze how users engage with LGBTQ+ news content. Our analyses focus both on positive and negative content. In particular, we construct a hope speech classifier that detects positive ( hope speech ), negative, neutral, and irrelevant content. Second, in consultation with a public health expert specializing on LGBTQ+ health, we conduct an annotation study with a balanced and diverse political representation and release a dataset of 3,750 instances with crowd-sourced labels and detailed annotator demographic information. Finally, beyond providing a vital resource for the LGBTQ+ community, our annotation study and subsequent in-the-wild assessments reveal (1) strong association between rater political beliefs and how they rate content relevant to a marginalized community, (2) models trained on individual political beliefs exhibit considerable in-the-wild disagreement, and (3) zero-shot large language models (LLMs) align more with liberal raters. Trigger Warning: this paper contains offensive material that some may find upsetting. 1 Introduction From crowdsourced storymapping projects sharing stories of love, loss, and a sense of belonging Kirby et al. (2021) to safe, anonymous spaces to seek resources McInroy et al. (2019) and dating platforms Blackwell et al. (2015) – the internet and modern technologies play a positive role in the health and well-being of the LGBTQ+ community in various ways. However, cyberbullying Abreu and Kenny (2018) , exposure to dehumanization through news media Mendelsohn et al. (2020) , and more recently, homophobic biases in large language models (LLMs) Dutta et al. (2024a) ; Dutta et al. (2025a) - are some of the modern technology perils the community grapples with. How do mainstream US cable news outlets cover the LGBTQ+ community and how do users interact with such news content? Via a substantial corpus of more than 130 million YouTube comments on more than 300K YouTube news videos uploaded on three major US cable news outlets, this paper investigates how LGBTQ+ discussions are situated in mainstream US political discourse. While prior literature has analyzed LGBTQ+ discourse in print media through the lens of a dehumanization framework Mendelsohn et al. (2020) , such analyses consulted a single source of news (The New York Times) and did not consider reader responses to LGBTQ+ news. In this work, we consider three major US cable news networks representing diverse political views (Fox News, CNN, and MSNBC) with a key goal to understand how users interact and engage with LGBTQ+ news content. In doing so, we not only focus on identifying negative discussions about the LGBTQ+ community, one of our key contributions is operationalizing the detection of hope speech championing the disadvantaged minority in the broader spirit of counter-speech literature Benesch (2014) ; Mathew et al. (2019) ; Palakodety et al. (2020b) ; Chakravarthi (2020) ; Hengle et al. (2024) . YouTube Comment Label This is why we have 500+ million guns in america. We need to start using them. If a man walked in my daughters bathroom, that would be his last day on earth. People arent gonna keep tolerating this. Theres about to be rainbow street justice youre gonna start seeing. When comes to my, or anyones kids, and i frankly dont care what your laws. Never did in the first place, but when it comes to my kid……ill wake up and chose violence EVERY time… Negative Hey if using the term f g is hate speech, then I stand guilty as charged. I hate what they stand for and I hate the corruption they’re bringing into society, and I especially hate that they are preying on the minds of children in an attempt to make them believe that their sick, perverted lifestyle is normal… Negative LISTEN PEOPLE: EVEN IF YOU DONT SUPPORT IT, YOU CAN RESPECT US AT LEAST. ALL WE WANT IS BASIC HUMAN DECENCY. LITERALLY JUST THAT. me and my girlfriend got spat at on a bus once, and then called homophobic slurs. we just want respect, being gay doesn’t change the fact that I’m a human… Positive Reading these comments are making me cry. I’m 13 and my Dad is transgender. 99% of Doctors accept it. Why are you one to argue with Science? Please, if you are a bigot, don’t reply to this comment. It will make me cry. I love my Dad, and you ignorant people who don’t accept science won’t change that…. Positive Table 1: Illustrative examples of positive and negative YouTube comments identified by our classifier in the wild. While hate speech detection and mitigation strategies have been extensively researched (see, e.g., Fortuna and Nunes, 2018 ), relatively little attention has gone into the opposite: hope speech . Defined as comments and posts that inspire optimism and diffuse hostility in online spaces Palakodety et al. (2020b) ; Chakravarthi (2020) , hope speech is crucial for groups who face a disproportionate amount of hate speech. In the current US cultural climate, where issues of race, gender, and sexuality are central to public debate Hartman (2019) , there is a need for tools that can detect both hope speech and negative content as listed in Table 1 and increase understanding of relevant online conversations. According to the FBI, one in five hate crimes in the US targets the LGBTQ+ community Federal Bureau of Investigation (2024) , and globally, homosexual activity remains punishable by death in six UN member states while only 37 countries recognize same-sex marriage Mendos and Rohaizad (2024) . Understanding and improving online discourse around LGBTQ+ issues is thus a vital step toward protecting a vulnerable community. Contributions. Our contributions are as follows. (1) In consultation with a health expert specializing in LGBTQ+ health for more than two decades, we curate a novel dataset of 3,750 instances with labels: neutral , irrelevant , positive (hope speech), and negative 1 1 1 The dataset is available at https://github.com/Social-Insights-Lab/LGBTQplus_HopeSpeech/ . . Each instance is labeled by three raters (one Republican, one Democrat, and one independent) ensuring diverse and balanced political perspectives. 98 raters self-identified as being part of the LGBTQ+ community. (2) We analyze the association between rater political beliefs and how they rate content relevant to the LGBTQ+ community. (3) We analyze the alignment of LLMs with political beliefs in connection with LGBTQ+ content and how rater biases affect model fine-tuning. (4) We provide novel insights into audience engagement patterns surrounding LGBTQ+ content, enhancing our understanding of LGBTQ+ discussions within mainstream US political discourse. To our knowledge, no such study exists at our scale. 2 Related Work Our work is closely related with the hope speech literature Palakodety et al. (2020b) ; Palakodety et al. (2020a) ; Chakravarthi (2020) ; Yoo et al. (2021) and the broader literature of counter speech Benesch (2014) ; Mathew et al. (2019) ; Hengle et al. (2024) ; Saha et al. (2022) ; Gupta et al. (2023) ; Saha et al. (2024) . Our work contributes to this literature through our unique focus on LGBTQ+ discourse

Read the complete text