Project Reflection

Rationale

My goal was to create a chatbot that provides users with different methods of coping with stress. The purpose of this project was to entertain, appease, and/or distract audiences through technology.

Risks, Challenges, and Opportunities

While the code free bot building method was rewarding in many ways, it also limited features that were crucial to my initial goal. First, the user was unable to actually “chat” with the bot. I noticed during trials that people naturally attempted to type their responses to the bot’s prompts. Second, Gupshup limited the number of options that I could propose in a single prompt. It also limited the number of characters for each response option. This was difficult to work around because more characters were needed in order for the conversation to sound like natural human interaction.

Despite these flaws, the code-free method allowed for success that I probably wouldn’t have achieved through coding myself. One bonus was the ability to link to resources and embed images. This enabled more creative conversational directions, like the red pill/blue pill ultimatum. It also allowed me to include visual aids for the mental health resources StressBot would provide to troubled users.

If I had to attempt this project again, I would try using Gupshup’s programming template. I think this would help me become more comfortable with coding and also allow me to create an actual chat bot without the character limits.

Resources

In addition to Gupshup, I used many different resources for this project. The goal was to provide ways to cope with stress and other related mental health issues in a digital context. I found a lot of great GIFs on Tumblr. I also found a ton of white noise/relaxing videos on YouTube that StressBot could link to. There were many informative websites that I gathered facts, tips and interactive exercises from that the bot provided as well.

Schedule

My initial schedule left me with a week of navigating/researching Gupshup and stress relief techniques, but this just didn’t happen as soon as I would’ve liked. I did manage to create an outline on time, but because I didn’t have enough resources or experience to thoroughly plan, I stopped relying on it. I researched the stress relief techniques and learned how to build the bot as I actually worked on it.

Outcomes

I’m happy with my end result; I attempted everything I could with the code-free method and I can continue adding more to it if I’d like. Because the decision to add and develop the “I’m OK” response wasn’t made until later on in the process, I feel like that half of the bot lacks variety. The “Not Good” response leads to a lot of the information and resources that I had in mind when I started this project.

I think I could have made the bot much more complex had I stuck to my schedule, but the direction I took ultimately led me to learn more about the function bots provide to humans. I discovered a lot of great online self-help resources while trying to build StressBot. I think that housing these resources in a personalized Messenger bot could benefit a lot of people who struggle with the problems I addressed. However, I don’t feel that what I created could be considered a “chat bot” because the user does not get a response by typing whatever they want.

Executive Summary

I created a Facebook Messenger bot that provides different methods of stress relief to entertain or help audiences. Through attempting this, I became accommodated with Gupshup, learned more about online mental health resources, and successfully published a bot to social media. Exploring Gupshup allowed me to realize that while there are many different ways to build a bot, each method has its own strengths and weaknesses. Because I fell a little off schedule, some areas of StressBot were not as complex as others. While the code free method had its limitations, it allowed me to quickly and efficiently build a bot that could potentially help people.

Send StressBot a message here.

 

Leave a Reply

Your email address will not be published. Required fields are marked *