MentorPal: Interactive Virtual Mentors Based on Real-Life STEM Professionals (bibtex)
by Nye, Benjamin D, Kaimakis, Nicholas J, Krishnamachari, Madhusudhan, Swartout, William, Campbell, Julia, Anderson, Clinton and Davis, Dan M
Abstract:
In an ideal world, all students could meet STEM role models as they explore different careers. However, events such as career fairs do not scale well: professionals have limited time and effective mentors are not readily available in all fields. The result is that students’ understanding is minimal about what professionals in STEM fields do every day, what education is needed, and even what STEM fields exist. Moreover, since in-person interactions rely on finding people engaged in current STEM careers, students may form career goals for stagnant fields rather than growing fields (e.g., projected workforce needs). To address this problem, we are designing a scalable tablet-based app that gives students the opportunity to converse with interactive recordings of real-life STEM professionals. These conversational virtual agents will emulate a question-and-answer session with STEM professionals who have Navy ties and who are engaging, enthusiastic, and effective mentors. These interactions will allow students to have a lifelike informational interview with a virtual agent whose responses are directly drawn from a specific real professional’s video-recorded interview. This work differs from prior research on career guides by capturing the experiences of a collection of unique mentors, which should be more authentic and engaging than a generic agent or resource which speaks only about the average experience. This paper will discuss the process of creating the first such virtual STEM mentor prototype, including the development of an extensive mentoring question bank (approximately 500 questions); key mentoring topics that intersect STEM, DoD, and civilian life; techniques for cost-effective recording of remote mentors; and the process of training and verifying a natural language dialogue model for answering and suggesting career questions. Finally, we conclude with implications, strengths, and drawbacks of virtualizing the experience of talking with specific mentors, from the perspectives of efficacy, scalability, and maintainability.
Reference:
MentorPal: Interactive Virtual Mentors Based on Real-Life STEM Professionals (Nye, Benjamin D, Kaimakis, Nicholas J, Krishnamachari, Madhusudhan, Swartout, William, Campbell, Julia, Anderson, Clinton and Davis, Dan M), In Proceedings of the Interservice/Industry Training, Simulation and Education Conference (I/ITSEC) 2017, a2z, Inc., 2017.
Bibtex Entry:
@inproceedings{nye_mentorpal:_2017,
	address = {Orlando, Florida},
	title = {{MentorPal}: {Interactive} {Virtual} {Mentors} {Based} on {Real}-{Life} {STEM} {Professionals}},
	url = {http://www.iitsecdocs.com/search},
	abstract = {In an ideal world, all students could meet STEM role models as they explore different careers. However, events such as career fairs do not scale well: professionals have limited time and effective mentors are not readily available in all fields. The result is that students’ understanding is minimal about what professionals in STEM fields do every day, what education is needed, and even what STEM fields exist. Moreover, since in-person interactions rely on finding people engaged in current STEM careers, students may form career goals for stagnant fields rather than growing fields (e.g., projected workforce needs). To address this problem, we are designing a scalable tablet-based app that gives students the opportunity to converse with interactive recordings of real-life STEM professionals. These conversational virtual agents will emulate a question-and-answer session with STEM professionals who have Navy ties and who are engaging, enthusiastic, and effective mentors. These interactions will allow students to have a lifelike informational interview with a virtual agent whose responses are directly drawn from a specific real professional’s video-recorded interview. This work differs from prior research on career guides by capturing the experiences of a collection of unique mentors, which should be more authentic and engaging than a generic agent or resource which speaks only about the average experience. This paper will discuss the process of creating the first such virtual STEM mentor prototype, including the development of an extensive mentoring question bank (approximately 500 questions); key mentoring topics that intersect STEM, DoD, and civilian life; techniques for cost-effective recording of remote mentors; and the process of training and verifying a natural language dialogue model for answering and suggesting career questions. Finally, we conclude with implications, strengths, and drawbacks of virtualizing the experience of talking with specific mentors, from the perspectives of efficacy, scalability, and maintainability.},
	booktitle = {Proceedings of the {Interservice}/{Industry} {Training}, {Simulation} and {Education} {Conference} ({I}/{ITSEC}) 2017},
	publisher = {a2z, Inc.},
	author = {Nye, Benjamin D and Kaimakis, Nicholas J and Krishnamachari, Madhusudhan and Swartout, William and Campbell, Julia and Anderson, Clinton and Davis, Dan M},
	month = nov,
	year = {2017},
	keywords = {Learning Sciences, MxR, UARC}
}
Powered by bibtexbrowser