Misinformation detection using multitask learning with mutual learning for novelty detection and emotion recognition

Kumari, Rina; Ashok, Nischal; Ghosal, Tirthankar; Ekbal, Asif

doi:10.1016/j.ipm.2021.102631

Cited by 44 publications

(22 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Individuals online often feel like they are in a perpetual arms race to keep people's attention, with incentives ranging from self-validation to profit. Natural human tendencies to seek out exciting new information [48] are easily co-opted by highly emotional [49] and potentially misleading information [50]. Some individuals thrive on posting novel, but inflammatory content to convey a morally righteous self-image, elicit positive feedback, or solidify ingroup status [51][52][53].…”

Section: The Internet: Echo Chambers Anonymity and The Audiencementioning

confidence: 99%

“…Some individuals thrive on posting novel, but inflammatory content to convey a morally righteous self-image, elicit positive feedback, or solidify ingroup status [51][52][53]. Although this provocative content gains traction easily, it is likely to misrepresent, dehumanize, or contain false information about an outgroup, discouraging productive CIC [49,50,54].…”

Section: The Internet: Echo Chambers Anonymity and The Audiencementioning

confidence: 99%

See 1 more Smart Citation

The Zoom solution: Promoting effective cross-ideological communication online

et al. 2022

View full text Add to dashboard Cite

The rise of ideological polarization in the U.S. over the past few decades has come with an increase in hostility on both sides of the political aisle. Although communication and compromise are hallmarks of a functioning society, research has shown that people overestimate the negative affect they will experience when viewing oppositional media, and it is likely that negative forecasts lead many to avoid cross-ideological communication (CIC) altogether. Additionally, a growing ideological geographic divide and online extremism fueled by social media audiences make engaging in CIC more difficult than ever. Here, we demonstrate that online video-chat platforms (i.e., Zoom) can be used to promote effective CIC among ideologically polarized individuals, as well as to better study CIC in a controlled setting. Participants (n = 122) had a face-to-face CIC over Zoom, either privately or publicly with a silent ingroup audience present. Participant forecasts about the interaction were largely inaccurate, with the actual conversation experience found to be more positive than anticipated. Additionally, the presence of an ingroup audience was associated with increased conflict. In both conditions, participants showed preliminary signs of attitude moderation, felt more favorable toward the outgroup, and felt more informed about the issue after the CIC. These results suggest that face-to-face CIC’s are generally positive and beneficial for polarized individuals, and that greater effects may be achieved through private conversations, as opposed to more public social media-like interactions. Future researchers studying ideological conflict may find success using similar Zoom paradigms to bring together ideologically diverse individuals in controlled lab settings.

show abstract

Section: The Internet: Echo Chambers Anonymity and The Audiencementioning

confidence: 99%

Section: The Internet: Echo Chambers Anonymity and The Audiencementioning

confidence: 99%

The Zoom solution: Promoting effective cross-ideological communication online

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Online audiences strongly in uence what content social media users decide to post, comment, or respond to. Individuals online often feel like they are in a perpetual arms race to keep people's attention, with incentives ranging from self-validation to pro t. Natural human tendencies to seek out exciting new information (42) are easily co-opted by highly emotional (43) and potentially misleading information (44). Some individuals thrive on posting novel, but in ammatory content to convey a morally righteous self-image, elicit positive feedback, or solidify ingroup status (45)(46)(47).…”

Section: Negative Forecasts and Avoidancementioning

confidence: 99%

“…Some individuals thrive on posting novel, but in ammatory content to convey a morally righteous self-image, elicit positive feedback, or solidify ingroup status (45)(46)(47). Although this provocative content gains traction easily, it is likely to misrepresent, dehumanize, or contain false information about an outgroup, discouraging productive CIC (43,44,48). Additionally, the presence of these "attitudinal ingroup" audiences may create a social pressure to espouse extreme opinions (49) to avoid "looking weak".…”

Section: Negative Forecasts and Avoidancementioning

confidence: 99%

The Zoom solution: Promoting effective cross-ideological communication online

Binnquist

Dolbier

Dieffenbach

et al. 2021

Preprint

View full text Add to dashboard Cite

The rise of ideological polarization in the U.S. over the past few decades has come with an increase in hostility on both sides of the political aisle. Although communication and compromise are hallmarks of a functioning society, research has shown that people overestimate the negative affect they will experience when viewing oppositional media, and it is likely that negative forecasts lead many to avoid cross-ideological communication (CIC) altogether. Additionally, a growing ideological geographic divide and online extremism fueled by social media audiences make engaging in CIC more difficult than ever. Here, we demonstrate that online video-chat platforms (i.e., Zoom) can be used to promote effective CIC among ideologically polarized individuals, as well as to better study CIC in a controlled setting. Participants (n = 122) had a face-to-face CIC over Zoom, either privately or publicly with a silent ingroup audience present. Participant forecasts about the interaction were largely inaccurate, with the actual conversation experience found to be more positive than anticipated. Additionally, the presence of an ingroup audience was associated with increased conflict. In both conditions, participants showed signs of attitude moderation, felt more favorable toward the outgroup, and felt more informed about the issue after the CIC. These results suggest that face-to-face CIC’s are generally positive and beneficial for partisans, and that greater effects may be achieved through private conversations, as opposed to more public social media interactions. Future researchers studying ideological conflict may find success using similar Zoom paradigms to bring together ideologically diverse individuals in controlled lab settings.

show abstract

“…With the onset of the COVID-19 outbreak, social media has served as an important tool for information generation, dissemination, and consumption, contributing to many high-quality COVID-19 related papers involving main aspects like infodemic, surveillance, and monitoring ( Shorten et al, 2021 ; Tsao et al, 2021 ). Infodemic papers are mainly about misinformation ( Agley & Xiao, 2021 ; Ahmed et al, 2020 ; Alaa et al, 2020 ), its detection ( Ayoub et al, 2021 ; Kouzy et al, 2020 ; Kumari et al, 2021 ), influence, exposure and spread ( Burel et al, 2020 , 2021 ; Hanyin et al, 2021 ; Obadimu et al, 2021 ; Tang, Fujimoto, et al, 2021 ). Surveillance and monitoring papers cover a wide range including assessing public sentiments ( Basiri et al, 2021 ; Blanco & Lourenço, 2022 ; Xuehua Han et al, 2020 ) like vaccine attitudes ( Aygun et al, 2021 ; Griffith et al, 2021 ; Lazarus et al, 2021 ), mental health ( Behl et al, 2021 ; Guntuku et al, 2020 ), and detecting or predicting COVID-19 trends ( Huang et al, 2021 ) and cases ( Shen et al, 2020 ).…”

Section: Introductionmentioning

confidence: 99%