NLP in CALL

ALT 15th Anniversary Conference

After watching the recorded presentations, join these authors for a live panel discussion on December 3, 2020 at 5:30 pm – 6:00 pm (CST). Moderator: Reza Neiriz

Presenter	Abstract
Hyejin Yang Full time researcher Chuang-Ang University, South Korea	AI chatbots as L2 conversation partners The rapid advance in Artificial Intelligence (AI), AI chatbot have been around for recent decades for different purposes. Existing commercial conversation chatbots such as Google Assistant (Google), Siri (Apple), or Alexa (Amazon) have been widely used for simple internet searches or for responding to individual users' inquiries about personal schedule, weather or news, and so forth. In the field of language education, there has been increasing interests to utilize chatbots as language learning partners. In this presentation, it will begin with introducing current AI chatbots that can be served as conversation partners for English learners. In addition, I, as a part of a research team on AI chatbots at a university in Korea, will also present my recent work on developing an AI chatbot and conducting several empirical research that aimed to find better ways to develop and to integrate chatbots into EFL classrooms. Video Recording
Stephanie Link Assistant Professor Oklahoma State University	Inspiring the next generation of genre-based automated writing evaluation research Automated writing evaluation (AWE) has evolved significantly in recent decades to meet the ever-changing needs of writers across educational settings. From what started as sentence-level automated feedback has now expanded to discourse- and/or genre-based feedback. Although AWE research continues to spark interest for general academic writing use (e.g., Ranalli & Yamashita, 2020; Link, Mehrzad, & Rahimi, 2020), there is a need for new directions in specialized ESP/EAP research and practice (Hyland & Wong, 2019) that can facilitate useful learning transfer (Loboto, 2006). Early genre-based AWE tools opened new possibilities for ESP/EAP, for example AntMover (Anthony, 2003) and IADE (Cotos, 2009), but with the advancement of NLP and AI approaches, more sophisticated tools and learning systems have emerged, such as the Research Writing Tutor (Cotos, 2014; Cotos & Pendar, 2016) and AcaWriter (Abel, Kitto, Knight, & Buckingham Shum, 2018). Nevertheless, there is a need for upward momentum that can spark new trends in tool creation and continue to bridge theory, research, and practice by carefully considering the needs of learners and ways in which software can most effectively contribute to learning (Anthony, 2019). This presentation will introduce a genre-based AWE tool called Wrangler for “rounding up” research writing resources. This web-based technology leverages the power of natural language processing to develop an intelligent tutoring system to enhance writing for publication. Development started with careful consideration of user experience design, including flow diagrams, wireframes, and 133 informal potential user interviews. A web analytics tool, Hotjar, was integrated into the interface design to track use and inform alpha-to-beta platform development. Wrangler has been integrated into writing for publication courses and workshops at Oklahoma State University; however, the team intends to expand Wrangler’s potential, transcend technological hurdles, and inspire a new generation of genre-based AWE research. Video Recording

Presenter

Abstract

Hyejin Yang

Full time researcher
Chuang-Ang University, South Korea

AI chatbots as L2 conversation partners

The rapid advance in Artificial Intelligence (AI), AI chatbot have been around for recent decades for different purposes. Existing commercial conversation chatbots such as Google Assistant (Google), Siri (Apple), or Alexa (Amazon) have been widely used for simple internet searches or for responding to individual users' inquiries about personal schedule, weather or news, and so forth. In the field of language education, there has been increasing interests to utilize chatbots as language learning partners.
In this presentation, it will begin with introducing current AI chatbots that can be served as conversation partners for English learners. In addition, I, as a part of a research team on AI chatbots at a university in Korea, will also present my recent work on developing an AI chatbot and conducting several empirical research that aimed to find better ways to develop and to integrate chatbots into EFL classrooms.

Video Recording

Stephanie Link

Assistant Professor
Oklahoma State University

Inspiring the next generation of genre-based automated writing evaluation research

Automated writing evaluation (AWE) has evolved significantly in recent decades to meet the ever-changing needs of writers across educational settings. From what started as sentence-level automated feedback has now expanded to discourse- and/or genre-based feedback. Although AWE research continues to spark interest for general academic writing use (e.g., Ranalli & Yamashita, 2020; Link, Mehrzad, & Rahimi, 2020), there is a need for new directions in specialized ESP/EAP research and practice (Hyland & Wong, 2019) that can facilitate useful learning transfer (Loboto, 2006). Early genre-based AWE tools opened new possibilities for ESP/EAP, for example AntMover (Anthony, 2003) and IADE (Cotos, 2009), but with the advancement of NLP and AI approaches, more sophisticated tools and learning systems have emerged, such as the Research Writing Tutor (Cotos, 2014; Cotos & Pendar, 2016) and AcaWriter (Abel, Kitto, Knight, & Buckingham Shum, 2018). Nevertheless, there is a need for upward momentum that can spark new trends in tool creation and continue to bridge theory, research, and practice by carefully considering the needs of learners and ways in which software can most effectively contribute to learning (Anthony, 2019). This presentation will introduce a genre-based AWE tool called Wrangler for “rounding up” research writing resources. This web-based technology leverages the power of natural language processing to develop an intelligent tutoring system to enhance writing for publication. Development started with careful consideration of user experience design, including flow diagrams, wireframes, and 133 informal potential user interviews. A web analytics tool, Hotjar, was integrated into the interface design to track use and inform alpha-to-beta platform development. Wrangler has been integrated into writing for publication courses and workshops at Oklahoma State University; however, the team intends to expand Wrangler’s potential, transcend technological hurdles, and inspire a new generation of genre-based AWE research.

Video Recording