A5058483396- AI in Service Interactions
- Mental Health via Writing
- Digital Mental Health Interventions
- Topic Modeling
- Natural Language Processing Techniques
- Mobile Health and mHealth Applications
- Sentiment Analysis and Opinion Mining
- Speech and dialogue systems
- Social Robot Interaction and HRI
- Misinformation and Its Impacts
- Artificial Intelligence in Games
- Language and cultural evolution
- Digital Communication and Language
Columbia University
2021-2025
Stanford University
2021
This study presents a pilot randomized controlled trial to assess the usability, feasibility, and initial efficacy of mobile app-based relational artificial intelligence (AI) chatbot (Exerbot) intervention for increasing physical activity behavior. The was conducted over 1-week period, during which participants were either converse with baseline without capacity (control group) or using social communication strategies. Objectively measured data collected smartphone pedometers. feasible in...
This paper investigates users’ speech rate adjustments during conversations with an Amazon Alexa socialbot in response to situational (in-lab vs. at-home) and communicative (ASR comprehension errors) factors. We collected user interaction studies measured at each turn the conversation baseline productions (collected prior interaction). Overall, we find that users slow their when talking bot, relative pre-interaction productions, consistent hyperarticulation. Speakers use even slower in-lab...
Using chatbots to make recommendations is increasingly popular. The design of recommendation has mainly been taking an information-centric approach by focusing on the recommended content per se. Limited attention how social connection and relational strategies, such as self-disclosure from a chatbot, may influence users' perception acceptance recommendation. In this work, we designed, implemented, evaluated chatbot capable performing three different levels self-disclosure: factual...
Gunrock 2.0 is built on top of with an emphasis user adaptation. combines various neural natural language understanding modules, including named entity detection, linking, and dialog act prediction, to improve understanding. Its management a hierarchical model that handles topics, such as movies, music, sports. The system-level manager can handle question acknowledgment, error handling, additional functions, making downstream modules much easier design implement. also adapts its topic...
Kai-Hui Liang, Sam Davidson, Xun Yuan, Shehan Panditharatne, Chun-Yen Chen, Ryan Shea, Derek Pham, Yinghua Tan, Erik Voss, Luke Fryer. Proceedings of the 18th Workshop on Innovative Use NLP for Building Educational Applications (BEA 2023). 2023.
Weixin Liang, Kai-Hui Zhou Yu. Proceedings of the 59th Annual Meeting Association for Computational Linguistics and 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.
Speech-based dialog systems primarily interact with users through their spoken responses. Understanding users' perception of, and subconscious behaviors toward, the system's speech are crucial for improving design. In current study, a voice chatbot designed having conversation in domain of music is used to test impact emotional expressiveness its text-to-speech (TTS) output. We parametrically manipulated degree via prosody lexical choice across conditions. two-pronged approach these effects...
Artificial intelligence chatbots are the vanguard in technology-based intervention to change people’s behavior. To develop chatbots, first step is understand natural language conversation strategies human conversation. This work introduces an dataset collected from a real-world physical activity program for women. We designed comprehensive annotation schemes four dimensions (domain, strategy, social exchange, and task-focused exchange) annotated subset of dialogs. built strategy classifier...
Using chatbots to deliver recommendations is increasingly popular. The design of recommendation has primarily been taking an information-centric approach by focusing on the recommended content per se. Limited attention how social connection and relational strategies, such as self-disclosure from a chatbot, may influence users' perception acceptance recommendation. In this work, we designed, implemented, evaluated chatbot capable performing three different levels self-disclosure: factual...
Open-domain dialog systems have a user-centric goal: to provide humans with an engaging conversation experience. User engagement is one of the most important metrics for evaluating open-domain systems, and could also be used as real-time feedback benefit policy learning. Existing work on detecting user disengagement typically requires hand-labeling many samples. We propose HERALD, efficient annotation framework that reframes training data process denoising problem. Specifically, instead...