It’s going to be okay: Measuring Access to Support in Online Communities

Wang, Zijian; Jurgens, David

doi:10.18653/v1/d18-1004

Cited by 48 publications

(37 citation statements)

References 72 publications

(49 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…1 Second, it varies in both content and tone. Third, users can develop strong identities on the site, which could facilitate user-level modeling, but these identities are generally pseudonymous, which is useful when studying charged social phenomena Wang and Jurgens, 2018). Fourth, the subreddit structure of the site creates opportunities to study the impact of condescension on community structure and norms (Buntain and Golbeck, 2014;Lin et al, 2017;Chandrasekharan et al, 2018).…”

Section: The Talkdown Corpusmentioning

confidence: 99%

TalkDown: A Corpus for Condescension Detection in Context

Wang

Potts

2019

Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conferen

Self Cite

View full text Add to dashboard Cite

Condescending language use is caustic; it can bring dialogues to an end and bifurcate communities. Thus, systems for condescension detection could have a large positive impact. A challenge here is that condescension is often impossible to detect from isolated utterances, as it depends on the discourse and social context. To address this, we present TALKDOWN, a new labeled dataset of condescending linguistic acts in context. We show that extending a language-only model with representations of the discourse improves performance, and we motivate techniques for dealing with the low rates of condescension overall. We also use our model to estimate condescension rates in various online communities and relate these differences to differing community norms.

show abstract

Section: The Talkdown Corpusmentioning

confidence: 99%

TalkDown: A Corpus for Condescension Detection in Context

Wang

Potts

2019

Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conferen

Self Cite

View full text Add to dashboard Cite

show abstract

“…There have been a few other computational studies that predict the effects of online self-presentation. Wang and Jurgens [37] examined how the presentation of gender on Reddit, StackExchange, and Wikipedia affects reactions of support. Bareket-Bojmel et al [5] examined how self-enhancing and self-derogating posting strategies affected responses to users with different goals on Facebook.…”

Section: Identity and Interactionmentioning

confidence: 99%

Phans, Stans and Cishets: Self-Presentation Effects on Content Propagation in Tumblr

Yoder

Shen

Wang

et al. 2020

12th ACM Conference on Web Science

View full text Add to dashboard Cite

Research on content propagation in social media has largely focused on features from the content of posts and the network structure of users. However, social media platforms are also spaces where users present their identities in particular ways. How do the ways users present themselves affect how content they produce is propagated? In this paper, we address this question with an empirical study of interaction and self-presentation data from Tumblr. We use a pairwise learning-to-rank framework to predict whether a given user will reblog (share) another user's post from features comparing selfpresented textual and visual identity information. We find evidence that alignment in identity presentation is associated with content propagation, as these features increase performance over a baseline of content features. Interpreting learned feature weights on selfpresented text identity labels, we find that users who present labels that match or indicate shared interests and values are generally more likely to propagate each other's content. CCS CONCEPTS • Human-centered computing → Collaborative and social computing; • Computing methodologies → Natural language processing.

show abstract

“…Other studies have looked at the effect of suicide prevention counselor training (Gould et al, 2013), identifying patterns of successful crisis hotline counselors (Althoff et al, 2016), automating counselor evaluation (Pérez-Rosas et al, 2017), and building a dashboard for crisis counselors (Dinakar et al, 2015). There is additional work to identify supportive and distressed behaviors and language in online forums (Balani and De Choudhury, 2015;De Choudhury and De, 2014;Wang and Jurgens, 2018) and support forum moderators (Hussain et al, 2015). Most similar to our study, was one study that showed the potential for an avatar system to help train medical doctors to deliver news to patients (Andrade et al, 2010).…”

Section: Related Workmentioning

confidence: 99%

Towards augmenting crisis counselor training by improving message retrieval

DeMasi¹,

Hearst²,

Recht³

2019

Proceedings of the Sixth Workshop on Computational Linguistics and Clinical Psychology

View full text Add to dashboard Cite

A fundamental challenge when training counselors is presenting novices with the opportunity to practice counseling distressed individuals without exacerbating a situation. Rather than replacing human empathy with an automated counselor, we propose simulating an individual in crisis so that human counselors in training can practice crisis counseling in a lowrisk environment. Towards this end, we collect a dataset of suicide prevention counselor roleplay transcripts and make initial steps towards constructing a CRISISbot for humans to counsel while in training. In this data-constrained setting, we evaluate the potential for message retrieval to construct a coherent chat agent in light of recent advances with text embedding methods. Our results show that embeddings can considerably improve retrieval approaches to make them competitive with generative models. By coherently retrieving messages, we can help counselors practice chatting in a low-risk environment.

show abstract

It’s going to be okay: Measuring Access to Support in Online Communities

Cited by 48 publications

References 72 publications

TalkDown: A Corpus for Condescension Detection in Context

TalkDown: A Corpus for Condescension Detection in Context

Phans, Stans and Cishets: Self-Presentation Effects on Content Propagation in Tumblr

Towards augmenting crisis counselor training by improving message retrieval

Contact Info

Product

Resources

About