Can machines talk? Comparison of Eliza with modern dialogue systems

Shah, Hanifa; Warwick, Kevin; Vallverdú, Jordi; Wu, Defeng

doi:10.1016/j.chb.2016.01.004

Cited by 96 publications

(42 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Despite these issues, both early (Weizenbaum, 1966) and very recent (Hasler, Tuchman, & Friedman, 2013;Hill, Ford, & Farreras, 2015) research with text-based Human-Chatbot interactions have consistently pointed to their potential benefits, particularly with regard to the motivation they seem to inspire in their users. Furthermore, research seeking to build directly on Weizenbaum' original efforts has suggested that current Chatbots have and continue to improve as many of them learn from their "roundthe-clock" web-based interaction with interested humans from around the globe (Shah, Warwick, Vallverdú, & Wu, 2016).…”

Section: Technology For Enhancing Interest and Learningmentioning

confidence: 99%

Stimulating and sustaining interest in a language course: An experimental comparison of Chatbot and Human task partners

Fryer

Ainley

Thompson

et al. 2017

Computers in Human Behavior

251

119

View full text Add to dashboard Cite

This is the accepted version prior to journal formatting. The official version can found at:https://authors.elsevier.com/a/1V8wS2f~UW0xXp-->FREE COPY the end of JULY 2017. Stimulating and sustaining interest in a language course: An experimental comparison of Chatbot and Human task partners AbstractNovel technology can be a powerful tool for enhancing students' interest in many learning domains. However, the sustainability and overall impact of such interest is unclear. This study tests the longer-term effects of technology on students' task and course interest. The experimental study was conducted with students in foreign language classes (n=122): a 12-week experimental trial that included pre-and post-course interest, and a sequence of task interest measures. Employing a counterbalanced design, at three week intervals students engaged in separate speaking tasks with each of a Human and "Chatbot" partner. Students' interest in successive tasks and in the course (pre-post), were used to assess differential partner effects and course interest development trajectories. Comparisons of task interest under different partner conditions over time indicated a significant drop in students' task interest with the Chatbot but not Human partner. After accounting for initial course interest, Structural Equation Modelling indicated that only task interest with the Human partner contributed to developing course interest. While Human partner task interest predicted future course interest, task interest under Chatbot partner conditions did not. Under Chatbot partner conditions there was a drop in task interest after the first task: a novelty effect. Implications for theory and practice are discussed.

show abstract

Section: Technology For Enhancing Interest and Learningmentioning

confidence: 99%

Stimulating and sustaining interest in a language course: An experimental comparison of Chatbot and Human task partners

Fryer

Ainley

Thompson

et al. 2017

Computers in Human Behavior

251

119

View full text Add to dashboard Cite

show abstract

“…This particular finding appears unique compared to other text-based interventions described in the literature but is consistent with NICH program emphases on building rapport and the importance of the human element in behavior change. In stark contrast with continued attempts to automate interventions, 21 the NICH program's text-messaging component remains truly personalized. This unique component, although more time consuming than automated delivery systems, may allow interventionists the needed flexibility to tailor text content to individual patient and family characteristics.…”

Section: Discussionmentioning

confidence: 99%

NICH at Its Best for Diabetes at Its Worst: Texting Teens and Their Caregivers for Better Outcomes

Wagner

Barry

Stoeckel

et al. 2017

J Diabetes Sci Technol

View full text Add to dashboard Cite

Background: There is growing evidence for the feasibility of text-based interventions for pediatric patients with type 1 diabetes (T1D). However, less is known regarding whether the use of personalized text messages with high-risk youth and their caregivers is associated with improvements in youth health. This study examines the use of diabetes-specific texts and associated health outcomes for participants of the Novel Interventions in Children's Healthcare (NICH) program. Methods:Text messages sent to youth with T1D and their caregivers during NICH intervention were coded regarding diabetes relevance and adherence-related content. Health outcome data (eg, HbA1c values, hospital admissions) prior to and following NICH program enrollment were collected.Results: Fewer than half (43%) of texts sent to patients and their caregivers were identified as being related to diabetes, and over 95% of diabetes-related texts were identified as adherence-related. Participants in the NICH program demonstrated a significant decrease in HbA1c values, t(23) = 2.78, p ≤ .05, and DKA-related hospital visits, t(24) = 2.78, p ≤ .01, during program involvement. Although no relationships were identified between patient-recipient text characteristics and health outcomes, the frequency and type of text messaging with caregivers was significantly associated with changes in health outcomes. Conclusions:This study represents the most extensive evaluation of diabetes-related SMS use and health outcomes for NICH participants to date. Findings demonstrate improvements in patient health during NICH program involvement. Implications include that sending frequent, personalized, and adherence-reinforcing texts to patients' caregivers may result in improved patient health, decreased utilization, and, potentially, associated reductions in health care costs.

show abstract

“…Chatbot applications have been around for a long time; the first well-known chatbot is Joseph Weizenbaum"s Eliza program developed in the early 1960s. Eliza facilitated the interaction between human and machine through a simple pattern matching and a template-based response mechanism to emulate the conversation [6,7].…”

Section: A Chatbots Applications and Usesmentioning

confidence: 99%

Generating and Analyzing Chatbot Responses using Natural Language Processing

Aleedy¹,

Shaiba²,

Bezbradica³

2019

IJACSA

View full text Add to dashboard Cite

Customer support has become one of the most important communication tools used by companies to provide before and after-sale services to customers. This includes communicating through websites, phones, and social media platforms such as Twitter. The connection becomes much faster and easier with the support of today's technologies. In the field of customer service, companies use virtual agents (Chatbot) to provide customer assistance through desktop interfaces. In this research, the main focus will be on the automatic generation of conversation "Chat" between a computer and a human by developing an interactive artificial intelligent agent through the use of natural language processing and deep learning techniques such as Long Short-Term Memory, Gated Recurrent Units and Convolution Neural Network to predict a suitable and automatic response to customers' queries. Based on the nature of this project, we need to apply sequence-to-sequence learning, which means mapping a sequence of words representing the query to another sequence of words representing the response. Moreover, computational techniques for learning, understanding, and producing human language content are needed. In order to achieve this goal, this paper discusses efforts towards data preparation. Then, explain the model design, generate responses, and apply evaluation metrics such as Bilingual Evaluation Understudy and cosine similarity. The experimental results on the three models are very promising, especially with Long Short-Term Memory and Gated Recurrent Units. They are useful in responses to emotional queries and can provide general, meaningful responses suitable for customer query. LSTM has been chosen to be the final model because it gets the best results in all evaluation metrics.

show abstract

Can machines talk? Comparison of Eliza with modern dialogue systems

Abstract: To find if current dialogue systems use the same, psychotherapist questioning technique as JosephPre-publication version.

Cited by 96 publications

References 20 publications

Stimulating and sustaining interest in a language course: An experimental comparison of Chatbot and Human task partners

Stimulating and sustaining interest in a language course: An experimental comparison of Chatbot and Human task partners

NICH at Its Best for Diabetes at Its Worst: Texting Teens and Their Caregivers for Better Outcomes

Generating and Analyzing Chatbot Responses using Natural Language Processing

Contact Info

Product

Resources

About