49714 Programming for Online Prototypes
· 9 members
A hands on introduction to building online products and services through code
For this project, I wanted to develop an application that was geared toward the healthcare sector because I am interested in the ways that design can improve the healthcare system. For a previous project, I had performed research on cancer patients and their caregivers. I found that many cancer caregivers experience depression and poor mental health because of the stresses that come with caring for a loved one. I decided that this was an opportunity to look at how conversational UIs could be used to improve the mental health of cancer caregivers.
One element I found from my research is that many caregivers would not express their feelings, especially if they were negative, such as anger, sadness or frustrations, to protect the patient. I thought that there was on opportunity to explore the act of releasing emotions and thoughts for this project. I was inspired by online patient/caregiver forums because they are a medium where people can currently release thoughts anonymously. I was also inspired by the 21 Days of Gratitude because I liked the integration of word and image to provide feedback to the user, which I decided I wanted to incorporate into the design of my application.
For this project, I knew I wanted to work with Google Assistant because I own a Google Home Mini and it would allow me to easily test my concept. I began with the intention of developing a VUI because I thought that speech would make it easier for the user to express their emotions in a natural way. I imagined that the user would be able to speak their thoughts to the Google Home Mini and then receive the visual feedback on their phone through the Google Assistant app. So I began developing the dialog with that intention, however, I decided to shift my focus to the text-based version through Google Assistant because I wanted to play with the rich messages, such as the cards and suggestion chips, and explore how they could help inform the user through the conversation.
I decided to focus on developing the workflow for the first time user because it would force me to think out the functions of the application because I would have to integrate the explanation of the functions in my conversation flow. I also took into account how the chatbot would respond if the user said yes vs no to a question and how the user would direct the conversational flow for the chatbot.
When I first began using dialog flow, I had run into the difficulty of dealing with yes or no answers. I found that a lot of my conversation relied on yes or no responses and that the chatbot would get confused and send the wrong response. I resolved this problem by using the follow-up intents. I also had the problem of anticipating what a user might respond with and I wanted to explore how I might direct the conversation. I worked with using the suggestion chips as a method to direct the conversation and make it easier for the user to understand how they might respond.
Because I wanted to develop this application to work using the Google Assistant, I worked with Actions on Google and Dialogflow, linking it to Heroku using Ruby code.
Release is a tool and community for cancer caregivers that provide them with a platform to express their thoughts and provide words of support and encouragement to other cancer caregivers. The two main functions of Release is the ability to add thoughts and the ability to add words of support to the Release community. I used the metaphor of a forest as my visual representation for Release. Each user has their own personal forest and there is a collective community forest. Each thought is represented by a leaf on the forest ground and each phrase of support is represented by a tree. When a user releases a thought, they receive a leaf which is added to their personal forest and the user has the option to share it with the community forest. When a user adds words of support, they create a tree which is added to the community forest. The user's personal forest is populated with support trees with words of support that other users have added to the community forest.
The main idea of this application is both to provide a platform to express emotions in a safe space, but also to demonstrate to others that it is ok and good for their mental health to release any emotions or thoughts that might come off negative. To show other cancer caregivers that they are not alone and that these thoughts they may be having that they may feel bad or guilty about happen to other caregivers and that it is ok.
I think that this project was fun and an interesting exercise because it made me think about how chatbots can be applicable to healthcare in non-traditional ways. It made me really think about how conversations are structured and the difficulty that goes into designing a natural conversation between a chatbot agent and a human.
I showed my prototype to a family member who was a cancer caregiver and she liked the idea of the concept and the metaphor of a forest and support trees. She felt that the chatbot's tone should include more positive language, she had noticed that I used "ok" a lot in the chatbot responses. I would like to continue working on this project and further develop the concept because I think that it has the potential to turn into a great little project for my portfolio.
A cancer caregiver tool and community where they can express their thoughts and provide words of support to other caregivers.
October 18th, 2018