Our group was inspired by the current climate to commit to something in regards to healthcare but realized very early on that COVID-oriented hacks were oversaturated and frequently done. Instead, why not combine multiple tracks together and come up with a project that both gives back to the community, functions as a viable life hack, and delves into healthcare, all around?
Recent laws being passed in Ottawa, Canada, made the sensitive topic of women's hygiene products a primary concern and, from there, IMPACT by FemHACKS was born.
The more we researched, the more we understood how very serious and charged the issue was and how much more attention needs to be paid to it. After all, from colloquial dialogue regarding "Auntie Flo'" to the more mundane slang of "that time of the month", even the avoidance of using the word "period" or "menstruation" has become sensitive censure.
The taboo that exists in today's society has only created more problems than it has solved. 4 out of 5 girls have to leave their schooling once a period hits, simply because they cannot afford feminine hygiene products to circumvent their natural bodily functions. Other diseases, such as unitary tract infections, cervical cancer, and more, also come into play when a woman's body cannot be taken care of by the products companies seem to deem a luxury, rather than a necessity!
The Solution? Time, Awareness, and the successful distribution of feminine hygiene products.
What It Does
IMPACT is a pseudo-chatbot app focused around the topic of feminine healthcare and hygiene. It receives questions from the user to respond to from a scraped-through database of questions and answers for women and by women that it's been trained on. Furthermore, it encourages people to clear all of their fears and doubts regarding menstrual hygiene and general related questions.
This hack is designed to educate people and spread awareness about menstrual hygiene. Our target audience is not only women but, too, the general public. We feel everyone, irrespective of gender should be aware of women's health and how best to help them stay healthy. People might not ask questions related to menstrual hygiene or discuss women's wellness with others due to the taboo that exists, after all.
With IMPACT, however, they can ask anything to our AI chatbot without fear of being judged or the stigma that comes from discussing something so "private" and "taboo" outside closed doors.
How We Built It
This project was built using Python and flask. The website deployed on the localhost, but the models were trained on the Google Colab. We gathered information in CSV files, which were later converted to JSON files and read through the model. The main website to scrap the data and train our model was "uestions.menstrupedia.com". To train our model, we used a reinforcement learning method. As the training structure user ask her question on the python terminal, and the chatbot will vectorize the problem with the help of Doc2vec and bring the best match question from its dataset. If the query was the same as the user's question, the user could confirm the question. Otherwise, the model faces a penalty for that and bring another query. In the end, we built a chatbot to integrate alongside the model for a better user experience overall with the help of Flask.
Challenges We Ran Into
The main challenges we ran into were deciding on what kind of product to make and learning how to use the different platforms and different algorithms to ultimately create a functioning hack. No data was readily available on platforms like Kaggle for menstrual hygiene, and most of our time was spent on research and scraping and extracting data from different websites.
On the other hand, time-zones were not as much an issue as expected, for the most part. However, errors in our output did delay us somewhat, but we finished surprisingly early.
Accomplishments That We're Proud Of
We're proud to have completed this project in the given timeframe and created something beneficial to the overall community, both to help in spreading awareness and to work with current-day news in ensuring new programs addressing feminine hygiene are successful.
What We Learned
We learned how to use different machine learning models in this project, and to delve deeper into topics and algorithms, we were not very comfortable with or used to.
What's Next for IMPACT
FemHACKS is, as of the moment, focused on a single application for Q&A. Side research in regards to location-tracking, as well as in referring users to non-profit organizations for feminine hygiene product distribution, were made. Still, we did not have time to fully implement both parts of the hack together.
In the future, we aim to keep track of local charities, churches, mosques, and various other non-profit locations where it is likely for the now-free feminine products to be distributed, as well as expand to new locales later on. Additionally, we plan to scrape data from more varied sources for better accuracy in our model's training to ensure we give our users the best end experience.
Basic necessities should be free for the people who cannot afford them, as basic necessities are not a privilege but, instead, a right. We believe it is shameful how some countries are still imposing taxes on menstrual hygiene products while most of the females within those countries cannot afford them. Further shameful, still, is the taboo that exists in a society that prevents women from getting the healthcare they deserve.
To combat this disparity, IMPACT is designed to ensure a successful program for the new laws being passed so that other cities, states, and countries may follow in their footsteps.