GDGinger Dragon

Ginger Dragon

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Ginger Dragon has a total of 11 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Sorting Pauses in Speech Based Dialog Systems: Effects Of Knowing Whether Pauses in Speech are Mid-Turn or at The End Of the Turn

    by Mohamed Jemmali, Ginger Dragon, Raelyn Mendoza, Daniel Stille, Patrick Poinar, Aditya Ojha, Kevin Baez, Sooyeon Jeong, Renu Balyan, Francisco Iacobelli

    Intelligent tutoring systems (ITS) can provide scalable and accessible education at scale. Generally, speech-based tutoring systems interpret the users' long pauses as a signal to take the floor. However, such behavior creates barriers for minority and low-literacy individuals who pause mid-sentence just to finish articulating their thoughts. For them, these systems inappropriately interrupt them, resulting in frustration. In this study, we present a bigram pause detection model that determines whether a pause is occurring in the middle or at the end of a sentence and implemented this model on an ITS (MODIFIED). We also compared the effect of this modification with a baseline ITS that interprets long pauses as the end of a turn (TRADITIONAL). Our results show that participants interacted longer with MODIFIED, and perceive it more human-like than the TRADITIONAL system. We argue that conversational systems need to be designed considering unique needs of their intended interlocutors.


  • Open Access

    Web-Based Information Seeking Behaviors of Low-Literacy Hispanic Survivors of Breast Cancer: Observational Pilot Study

    by Francisco Iacobelli, Ginger Dragon, Giselle Mazur, Judith Guitelman

    Background: Internet searching is a useful tool for seeking health information and one that can benefit low-literacy populations. However, low-literacy Hispanic survivors of breast cancer do not normally search for health information on the web. For them, the process of searching can be frustrating, as frequent mistakes while typing can result in misleading search results lists. Searches using voice (dictation) are preferred by this population; however, even if an appropriate result list is displayed, low-literacy Hispanic women may be challenged in their ability to fully understand any individual article from that list because of the complexity of the writing. Objective: This observational study aims to explore and describe web-based search behaviors of Hispanic survivors of breast cancer by themselves and with their caregivers, as well as to describe the challenges they face when processing health information on the web. Methods: We recruited 7 Hispanic female survivors of breast cancer. They had the option to bring a caregiver. Of the 7 women, 3 (43%) did, totaling 10 women. We administered the Health LiTT health literacy test, a demographic survey, and a breast cancer knowledge assessment. Next, we trained the participants to search on the web with either a keyboard or via voice. Then, they had to find information about 3 guided queries and 1 free-form query related to breast cancer. Participants were allowed to search in English or in Spanish. We video and audio recorded the computer activity of all participants and analyzed it. Results: We found web articles to be written for a grade level of 11.33 in English and 7.15 in Spanish. We also found that most participants preferred searching using voice but struggled with this modality. Pausing while searching via voice resulted in incomplete search queries, as it confused the search engine. At other times, background noises were detected and included in the search. We also found that participants formulated overly general queries to broaden the results list hoping to find more specific information. In addition, several participants considered their queries satisfied based on information from the snippets on the result lists alone. Finally, participants who spent more time reviewing articles scored higher on the health literacy test. Conclusions: Despite the problems of searching using speech, we found a preference for this modality, which suggests a need to avoid potential errors that could appear in written queries. We also found the use of general questions to increase the chances of answers to more specific concerns. Understanding search behaviors and information evaluation strategies for low-literacy Hispanic women survivors of breast cancer is fundamental to designing useful search interfaces that yield relevant and reliable information on the web.


  • Open Access

    Web-Based Information Seeking Behaviors of Low-Literacy Hispanic Survivors of Breast Cancer: Observational Pilot Study (Preprint)

    by Francisco Iacobelli, Ginger Dragon, Giselle Mazur, Judith Guitelman

    BACKGROUND Internet searching is a useful tool for seeking health information and one that can benefit low-literacy populations. However, low-literacy Hispanic survivors of breast cancer do not normally search for health information on the web. For them, the process of searching can be frustrating, as frequent mistakes while typing can result in misleading search results lists. Searches using voice (dictation) are preferred by this population; however, even if an appropriate result list is displayed, low-literacy Hispanic women may be challenged in their ability to fully understand any individual article from that list because of the complexity of the writing. OBJECTIVE This observational study aims to explore and describe web-based search behaviors of Hispanic survivors of breast cancer by themselves and with their caregivers, as well as to describe the challenges they face when processing health information on the web. METHODS We recruited 7 Hispanic female survivors of breast cancer. They had the option to bring a caregiver. Of the 7 women, 3 (43%) did, totaling 10 women. We administered the Health LiTT health literacy test, a demographic survey, and a breast cancer knowledge assessment. Next, we trained the participants to search on the web with either a keyboard or via voice. Then, they had to find information about 3 guided queries and 1 free-form query related to breast cancer. Participants were allowed to search in English or in Spanish. We video and audio recorded the computer activity of all participants and analyzed it. RESULTS We found web articles to be written for a grade level of 11.33 in English and 7.15 in Spanish. We also found that most participants preferred searching using voice but struggled with this modality. Pausing while searching via voice resulted in incomplete search queries, as it confused the search engine. At other times, background noises were detected and included in the search. We also found that participants formulated overly general queries to broaden the results list hoping to find more specific information. In addition, several participants considered their queries satisfied based on information from the snippets on the result lists alone. Finally, participants who spent more time reviewing articles scored higher on the health literacy test. CONCLUSIONS Despite the problems of searching using speech, we found a preference for this modality, which suggests a need to avoid potential errors that could appear in written queries. We also found the use of general questions to increase the chances of answers to more specific concerns. Understanding search behaviors and information evaluation strategies for low-literacy Hispanic women survivors of breast cancer is fundamental to designing useful search interfaces that yield relevant and reliable information on the web. CLINICALTRIAL