Lu Xiao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Lu Xiao has a total of 142 co-authors. Here are the top co-authors by number of shared publications:
- John Millar Carroll
College of Information Sciences and Technology, The Pennsylvania State University, University Park, PA, USA
15 shared publications Mary Beth Rosson
College of Information Sciences and Technology and Center for Human-Computer Interaction, The Pennsylvania State University, University Park, 16802 U.S.A.
9 shared publicationsTaraneh Khazaei
Western University, London, Canada
7 shared publications- Nicole Askin
WRHA Virtual Library, University of Manitoba, Winnipeg, MB R3E 3P5, Canada
4 shared publications Umer Farooq
4 shared publications- Robert E. Mercer
The University of Western Ontario
3 shared publications
Latest Publications
- 5 April 2024
- Published by Emerald Publishing in Online Information Review
- Vol. 48 (6) , 1197-1215
Persuading others in different communication media: appeals to logic, authority and emotion
Purpose: Scholars of persuasion have long made a distinction between appeals to logic, emotion and authority- logos, ethos and pathos- but ideas developed to account for live face-to-face conversation processes must also be tested in new media. We aimed to test the effectiveness of these three strategies in one-to-one chats through different communication media.Design/methodology/approach: With a 3 × 3 × 2 between-subject factorial design, we tested these three strategies in one-to-one chats (female–female or male–male pairs) through three communication media: face-to-face, Skype video or Skype text. The persuasion scenario was adapted from prior studies in which students were presented with the idea of requiring a comprehensive exam as part of their degree. The participants were all undergraduate students of a major university in USA.Findings: Our results showed trivial differences between female–female and male–male conditions. The logos appeal worked best overall in persuading the participants to change their reported attitudes. Additionally, the explanations provided by the participants for their own opinions were most like the persuasion scripts in the logos condition compared to the other two appeal conditions. Separately, participants indicated some disapproval of the pathos appeal in the text-based chat condition, although this did not seem to make a difference in terms of actual attitude change.Research limitations/implications: One major limitation of our study is that our subjects are college students and therefore are not representative of Internet users in general. Future research should test these three types of persuasion strategies on people of diverse backgrounds. For example, while logos seems to be most effective strategy in persuading college students (at least in our study), pathos or ethos may be more effective when one attempts to persuade people of different backgrounds.Practical implications: Although it is enough for a statistical test, our sample size is still relatively small due to constraints on time, personnel and funding. We also recognize that it is challenging both conceptually and empirically to compare the effectiveness of three persuasion strategies separately.Social implications: Our findings suggest it is helpful to use fact-checking tools to combat disinformation in cases where users may not have sufficient domain knowledge or may not realize the need to identify or examine the given information. Additionally, it may require more effort to negate the impact of the disinformation spread than correcting the information, as some users may not only believe false information but also may start to reason in ways similar to those presented in the disinformation messages.Originality/value: Past studies on online persuasion have limitedly examined whether and how communication media and persuasion strategies interact in one-to-one persuasion sessions. Our experiment makes an attempt to close this gap by examining the persuasion process and outcome in three different communication media and with three different persuasion strategies.
- 5 April 2024
- Published by Emerald Publishing in Online Information Review
- Vol. 48 (6) , 1197-1215
Persuading others in different communication media: appeals to logic, authority and emotion
Purpose: Scholars of persuasion have long made a distinction between appeals to logic, emotion and authority- logos, ethos and pathos- but ideas developed to account for live face-to-face conversation processes must also be tested in new media. We aimed to test the effectiveness of these three strategies in one-to-one chats through different communication media.Design/methodology/approach: With a 3 × 3 × 2 between-subject factorial design, we tested these three strategies in one-to-one chats (female–female or male–male pairs) through three communication media: face-to-face, Skype video or Skype text. The persuasion scenario was adapted from prior studies in which students were presented with the idea of requiring a comprehensive exam as part of their degree. The participants were all undergraduate students of a major university in USA.Findings: Our results showed trivial differences between female–female and male–male conditions. The logos appeal worked best overall in persuading the participants to change their reported attitudes. Additionally, the explanations provided by the participants for their own opinions were most like the persuasion scripts in the logos condition compared to the other two appeal conditions. Separately, participants indicated some disapproval of the pathos appeal in the text-based chat condition, although this did not seem to make a difference in terms of actual attitude change.Research limitations/implications: One major limitation of our study is that our subjects are college students and therefore are not representative of Internet users in general. Future research should test these three types of persuasion strategies on people of diverse backgrounds. For example, while logos seems to be most effective strategy in persuading college students (at least in our study), pathos or ethos may be more effective when one attempts to persuade people of different backgrounds.Practical implications: Although it is enough for a statistical test, our sample size is still relatively small due to constraints on time, personnel and funding. We also recognize that it is challenging both conceptually and empirically to compare the effectiveness of three persuasion strategies separately.Social implications: Our findings suggest it is helpful to use fact-checking tools to combat disinformation in cases where users may not have sufficient domain knowledge or may not realize the need to identify or examine the given information. Additionally, it may require more effort to negate the impact of the disinformation spread than correcting the information, as some users may not only believe false information but also may start to reason in ways similar to those presented in the disinformation messages.Originality/value: Past studies on online persuasion have limitedly examined whether and how communication media and persuasion strategies interact in one-to-one persuasion sessions. Our experiment makes an attempt to close this gap by examining the persuasion process and outcome in three different communication media and with three different persuasion strategies.
- 14 February 2024
- Published by Springer Nature
- p. 575-585
Toward Detection of Fake News Using Sentiment Analysis for Albanian News Articles
- 14 February 2024
- Published by Springer Nature
- p. 575-585
Toward Detection of Fake News Using Sentiment Analysis for Albanian News Articles
- Open Access
- 30 April 2023
- Published by Association for Computing Machinery (ACM)
- p. 128-135
Multi-emotion Recognition Using Multi-EmoBERT and Emotion Analysis in Fake News
Emotion recognition techniques are increasingly applied in fake news veracity or stance detection. While multiple co-existing emotions tend to co-occur in a single news article, most existing fake news detection has only leveraged single-label emotion recognition mechanisms. In addition, the relationship between the emotion of an article and its stance has not been sufficiently explored. To address these research gaps, we have developed and applied a multi-label emotion recognition tool called Multi-EmoBERT in fake news datasets. The tool delivers state-of-the-art performance on SemEval2018 Task 1. We apply the tool to identify emotions in several fake news datasets and examine the relationships between veracity/stance and emotion. Our work demonstrates the potential for predicting multiple co-existing emotions for fake news and implications against fake news spread.
- 30 April 2023
- Published by Association for Computing Machinery (ACM)
- p. 128-135
Open AccessMulti-emotion Recognition Using Multi-EmoBERT and Emotion Analysis in Fake News
Emotion recognition techniques are increasingly applied in fake news veracity or stance detection. While multiple co-existing emotions tend to co-occur in a single news article, most existing fake news detection has only leveraged single-label emotion recognition mechanisms. In addition, the relationship between the emotion of an article and its stance has not been sufficiently explored. To address these research gaps, we have developed and applied a multi-label emotion recognition tool called Multi-EmoBERT in fake news datasets. The tool delivers state-of-the-art performance on SemEval2018 Task 1. We apply the tool to identify emotions in several fake news datasets and examine the relationships between veracity/stance and emotion. Our work demonstrates the potential for predicting multiple co-existing emotions for fake news and implications against fake news spread.
- 14 October 2022
- Published by Wiley in Proceedings of the Association for Information Science and Technology
- Vol. 59 (1) , 840-842
Persuasion in Audio‐based Social Media: A Clubhouse Interview Study
Clubhouse is an audio‐based social media platform launched in March 2020. In this platform, users create chat rooms for discussions around a topic of their choice. Organizers of these discussions can decide which listener in the room to be given the right to speak. To explore their experiences of persuading others and of being persuaded in Clubhouse discussions, we interviewed 16 Clubhouse users. One's credibility is important to the perceived persuasiveness of one's argument. In Clubhouse, one may establish their credibility through their profile pages where they introduce themselves such as their backgrounds, experiences, affiliations, passions, etc. Hence, our interviews also probed whether speakers' profile pages are used to check their credibility.
- 14 October 2022
- Published by Wiley in Proceedings of the Association for Information Science and Technology
- Vol. 59 (1) , 840-842
Persuasion in Audio‐based Social Media: A Clubhouse Interview Study
Clubhouse is an audio‐based social media platform launched in March 2020. In this platform, users create chat rooms for discussions around a topic of their choice. Organizers of these discussions can decide which listener in the room to be given the right to speak. To explore their experiences of persuading others and of being persuaded in Clubhouse discussions, we interviewed 16 Clubhouse users. One's credibility is important to the perceived persuasiveness of one's argument. In Clubhouse, one may establish their credibility through their profile pages where they introduce themselves such as their backgrounds, experiences, affiliations, passions, etc. Hence, our interviews also probed whether speakers' profile pages are used to check their credibility.
- 7 July 2022
- Published by Springer Nature
- p. 787-799
Language Use and Susceptibility in Online Conversation
- 7 July 2022
- Published by Springer Nature
- p. 787-799
Language Use and Susceptibility in Online Conversation
- 7 July 2022
- Published by Springer Nature
- p. 800-813
How Does the Thread Level of a Comment Affect its Perceived Persuasiveness? A Reddit Study
- 7 July 2022
- Published by Springer Nature
- p. 800-813
How Does the Thread Level of a Comment Affect its Perceived Persuasiveness? A Reddit Study
- Open Access
- 28 January 2022
- Published by Open Library of the Humanities in Digital Studies / Le Champ Numérique
- Vol. 12 (1)
Tension Analysis in Survivor Interviews: A Computational Approach
This study aims to develop computational techniques to analyze and identify points of tensions in interviews with survivors of the 1994 Rwandan genocide. Oral history interviews are a dialogical source composed of questions and answers, producing a conversational narrative. Yet survivor testimony is often approached as though the questions did not exist. This article examines a digital tool that helps us visualize and better understand the underlying interview dynamic that is the heart of oral history and qualitative research more generally. Our tension detection tool identifies those moments in the interview when the interviewer and interviewee are trying to pull the conversation in different directions. This is part of the natural give-and-take of the interview. Hedging, deflection, hesitation, and boosting are all critical components of this interviewer-interviewee tension. By making the interview dynamic central to our analysis, we aim to better understand how the interview dynamic shapes what is being said and what is left unsaid. In this study, we address key components of interview tension and propose a natural language processing model that can efficiently incorporate these components in text-based oral history interviews to identify tension points. With experiments on an annotated transcript, we verify the efficacy of our model. This model provides a framework that can be utilized in future research on the dialogic of the interview.Cette étude vise à développer des techniques computationnelles pour analyser et identifier des points de tensions dans des interviews avec des survivants du génocide rwandais en 1994. Les interviews d’histoire orale sont une source dialogique composée de questions et de réponses, ce qui produit une narration conversationnelle. Cependant, le témoignage de survivant est souvent traité comme si les questions n’existaient pas. Cet article examine un outil numérique qui nous aide à visualiser et à mieux comprendre la dynamique d’interview sous-jacente qui est au cœur de l’histoire orale et, plus généralement, au cœur de la recherche qualitative. Notre outil détecteur de tensions identifie ces moments dans l’interview lorsque l’intervieweur et l’interviewé sont en train d’essayer de guider la conversation dans des directions différentes. Cela fait partie de l’interaction bidirectionnelle naturelle d’une interview. Le non-engagement, le détournement, l’hésitation et l’exagération sont tous des composants essentiels dans la tension intervieweur-interviewé. En mettant la dynamique d’interview au centre de notre analyse, nous aspirons à mieux comprendre comment la dynamique d’interview structure ce qui est dit et ce qui n’est pas dit. Dans cette étude, nous abordons des composants clés de la tension d’interview et proposons un modèle de traitement de langue naturelle qui peut incorporer de façon efficace ces composants dans des interviews d’histoire orale numérisées afin d’identifier des points de tensions. Avec des expériences sur un transcrit annoté, nous vérifions l’efficacité de notre modèle. Ce modèle fournit un cadre à utiliser dans de futures recherches sur la dialogique de l’interview.
- 28 January 2022
- Published by Open Library of the Humanities in Digital Studies / Le Champ Numérique
- Vol. 12 (1)
Open AccessTension Analysis in Survivor Interviews: A Computational Approach
This study aims to develop computational techniques to analyze and identify points of tensions in interviews with survivors of the 1994 Rwandan genocide. Oral history interviews are a dialogical source composed of questions and answers, producing a conversational narrative. Yet survivor testimony is often approached as though the questions did not exist. This article examines a digital tool that helps us visualize and better understand the underlying interview dynamic that is the heart of oral history and qualitative research more generally. Our tension detection tool identifies those moments in the interview when the interviewer and interviewee are trying to pull the conversation in different directions. This is part of the natural give-and-take of the interview. Hedging, deflection, hesitation, and boosting are all critical components of this interviewer-interviewee tension. By making the interview dynamic central to our analysis, we aim to better understand how the interview dynamic shapes what is being said and what is left unsaid. In this study, we address key components of interview tension and propose a natural language processing model that can efficiently incorporate these components in text-based oral history interviews to identify tension points. With experiments on an annotated transcript, we verify the efficacy of our model. This model provides a framework that can be utilized in future research on the dialogic of the interview.Cette étude vise à développer des techniques computationnelles pour analyser et identifier des points de tensions dans des interviews avec des survivants du génocide rwandais en 1994. Les interviews d’histoire orale sont une source dialogique composée de questions et de réponses, ce qui produit une narration conversationnelle. Cependant, le témoignage de survivant est souvent traité comme si les questions n’existaient pas. Cet article examine un outil numérique qui nous aide à visualiser et à mieux comprendre la dynamique d’interview sous-jacente qui est au cœur de l’histoire orale et, plus généralement, au cœur de la recherche qualitative. Notre outil détecteur de tensions identifie ces moments dans l’interview lorsque l’intervieweur et l’interviewé sont en train d’essayer de guider la conversation dans des directions différentes. Cela fait partie de l’interaction bidirectionnelle naturelle d’une interview. Le non-engagement, le détournement, l’hésitation et l’exagération sont tous des composants essentiels dans la tension intervieweur-interviewé. En mettant la dynamique d’interview au centre de notre analyse, nous aspirons à mieux comprendre comment la dynamique d’interview structure ce qui est dit et ce qui n’est pas dit. Dans cette étude, nous abordons des composants clés de la tension d’interview et proposons un modèle de traitement de langue naturelle qui peut incorporer de façon efficace ces composants dans des interviews d’histoire orale numérisées afin d’identifier des points de tensions. Avec des expériences sur un transcrit annoté, nous vérifions l’efficacité de notre modèle. Ce modèle fournit un cadre à utiliser dans de futures recherches sur la dialogique de l’interview.
- 4 May 2021
- Published by Emerald Publishing in Online Information Review
- Vol. 45 (7) , 1208-1226
Reasoning in social media: insights from Reddit “Change My View” submissions
Purpose: Complex social interactions such as argumentation and persuasion are increasingly common in online communications. To better understand these interactions and their impacts on people and on the society, it is important for the authors to understand how people reason online such as when they need to justify themselves or convince others with their perspectives. Reasoning in online discussions is expectedly to be different from doing so offline, as one often communicates with others anonymously and asynchronously in such contexts (e.g. Reddit discussions). The purpose of this paper is to investigate people's reasoning behavior in online environments focusing on how they justify their perspectives.Design/methodology/approach: In this study, the authors examined how a subreddit Change My View (CMV) users offer their opinions and justify them through the lens of argumentation and reasoning. The authors annotated, 330 Reddit discussion original posts (OPs) to identify claims, rationales and supports for reasoning, i.e. personal experiences, definitions, domain expertise and external sources. The authors investigated the correlations among the occurrences of these supporting statements and whether they are related to the topics of the posts.Findings: The findings suggest that if people mention their domain expertise, they tend to provide related personal experiences as well. Additionally, if the participants talk about the topic of domestic politics, they tend to utilize their personal experiences.Research limitations/implications: Further research may be conducted to help system designers. System designers (e.g. online debate systems, collective decision-making systems, etc.) may benefit from the findings to design systems by considering commonly used supporting statements, which may enhance people's reasoning and argumentation processes. The sample size is a small sample. The authors acknowledge that the small sample size of the study may limit the generalizability of the findings; however, it is still acceptable compared to the existing literature. One future study could be annotating a larger dataset to further probe the use of supporting statements in online reasoning.Practical implications: The authors' findings might be useful to understand how Reddit users are justifying their opinions as the reflection of their reasoning processes. In order to contribute further research in argumentation and reasoning in online platforms, the authors make the annotated dataset publicly available.Originality/value: To best of the authors' knowledge, this study was one of a few studies whose purpose is to understand Reddit CMV users' reasoning processes. To understand how online users offer their reasons while providing their ideas is important to have effective communication processes and to improve online discussion experiences which are very common in today's digital era.Peer review: The peer review history for this article is available at: https://publons.com/publon/10.1108/OIR-08-2020-0330
- 4 May 2021
- Published by Emerald Publishing in Online Information Review
- Vol. 45 (7) , 1208-1226
Reasoning in social media: insights from Reddit “Change My View” submissions
Purpose: Complex social interactions such as argumentation and persuasion are increasingly common in online communications. To better understand these interactions and their impacts on people and on the society, it is important for the authors to understand how people reason online such as when they need to justify themselves or convince others with their perspectives. Reasoning in online discussions is expectedly to be different from doing so offline, as one often communicates with others anonymously and asynchronously in such contexts (e.g. Reddit discussions). The purpose of this paper is to investigate people's reasoning behavior in online environments focusing on how they justify their perspectives.Design/methodology/approach: In this study, the authors examined how a subreddit Change My View (CMV) users offer their opinions and justify them through the lens of argumentation and reasoning. The authors annotated, 330 Reddit discussion original posts (OPs) to identify claims, rationales and supports for reasoning, i.e. personal experiences, definitions, domain expertise and external sources. The authors investigated the correlations among the occurrences of these supporting statements and whether they are related to the topics of the posts.Findings: The findings suggest that if people mention their domain expertise, they tend to provide related personal experiences as well. Additionally, if the participants talk about the topic of domestic politics, they tend to utilize their personal experiences.Research limitations/implications: Further research may be conducted to help system designers. System designers (e.g. online debate systems, collective decision-making systems, etc.) may benefit from the findings to design systems by considering commonly used supporting statements, which may enhance people's reasoning and argumentation processes. The sample size is a small sample. The authors acknowledge that the small sample size of the study may limit the generalizability of the findings; however, it is still acceptable compared to the existing literature. One future study could be annotating a larger dataset to further probe the use of supporting statements in online reasoning.Practical implications: The authors' findings might be useful to understand how Reddit users are justifying their opinions as the reflection of their reasoning processes. In order to contribute further research in argumentation and reasoning in online platforms, the authors make the annotated dataset publicly available.Originality/value: To best of the authors' knowledge, this study was one of a few studies whose purpose is to understand Reddit CMV users' reasoning processes. To understand how online users offer their reasons while providing their ideas is important to have effective communication processes and to improve online discussion experiences which are very common in today's digital era.Peer review: The peer review history for this article is available at: https://publons.com/publon/10.1108/OIR-08-2020-0330
- Open Access
- 22 October 2020
- Published by Wiley in Proceedings of the Association for Information Science and Technology
- Vol. 57 (1)
Preparing for the future of work: Building an informal learning community about intelligent technologies in poor neighborhoods
As intelligent machines continue to shape current and future work, low‐income families are expected to be the most negatively affected. To sustainably help low‐income families be better prepared for a future of working with intelligent machines, we are building an informal learning community in a poor neighborhood where we engage children and parents through a robotics club. We report our community building progress and the interviews we conducted with the children and their parents.
- 22 October 2020
- Published by Wiley in Proceedings of the Association for Information Science and Technology
- Vol. 57 (1)
Open AccessPreparing for the future of work: Building an informal learning community about intelligent technologies in poor neighborhoods
As intelligent machines continue to shape current and future work, low‐income families are expected to be the most negatively affected. To sustainably help low‐income families be better prepared for a future of working with intelligent machines, we are building an informal learning community in a poor neighborhood where we engage children and parents through a robotics club. We report our community building progress and the interviews we conducted with the children and their parents.
- Open Access
- 22 October 2020
- Published by Wiley in Proceedings of the Association for Information Science and Technology
- Vol. 57 (1)
Emotions in online debates: Tales from 4Forums and ConvinceMe
It is increasingly common for people to debate over various topics through online debate forums. While it has been shown that participants' emotional states affect debate processes and outcomes, it is unknown how different types of emotions are represented in online debates and what correlations exist between the emotions and other aspects of the debates such as their debate topic. We conduct a large‐scale analysis of the emotions in two online debate forums, namely, 4Forums and ConvinceMe. Specifically, we first develop an emotion recognition algorithm that uses multiple channels BLSTM with a feedforward attention mechanism, which outperforms the state‐of‐the‐art emotion recognition algorithm. Next, we label the emotions of each comment in the selected 4Forums and ConvinceMe discussions and analyze various aspects of the emotion's influence in the online debates. We observe that certain types of emotions are more likely dependent on the debate topic, and the prevalence of different emotions is independent of the individual discussions. We also observe emotion contagion between a comment and the immediately previous comment. We investigate the emotions of different types of respondents are less likely to express joy when they disagree and more likely to express disgust when they attack or disrespect to others.
- 22 October 2020
- Published by Wiley in Proceedings of the Association for Information Science and Technology
- Vol. 57 (1)
Open AccessEmotions in online debates: Tales from 4Forums and ConvinceMe
It is increasingly common for people to debate over various topics through online debate forums. While it has been shown that participants' emotional states affect debate processes and outcomes, it is unknown how different types of emotions are represented in online debates and what correlations exist between the emotions and other aspects of the debates such as their debate topic. We conduct a large‐scale analysis of the emotions in two online debate forums, namely, 4Forums and ConvinceMe. Specifically, we first develop an emotion recognition algorithm that uses multiple channels BLSTM with a feedforward attention mechanism, which outperforms the state‐of‐the‐art emotion recognition algorithm. Next, we label the emotions of each comment in the selected 4Forums and ConvinceMe discussions and analyze various aspects of the emotion's influence in the online debates. We observe that certain types of emotions are more likely dependent on the debate topic, and the prevalence of different emotions is independent of the individual discussions. We also observe emotion contagion between a comment and the immediately previous comment. We investigate the emotions of different types of respondents are less likely to express joy when they disagree and more likely to express disgust when they attack or disrespect to others.