Equify

Children from a young age are exposed to books to advance their learning. Currently, the books children read contain implicit gender and racial biases that negatively influence their perceptions as adults. With the importance of literature in all education, we are dedicated to reducing racial bias in books, specifically e-books and children’s books. Our project aims to create more equitable education early on to prevent those biases from harming children’s perceptions of themselves and others. Utilizing machine learning, we 1) replace the skin colors of characters with a more diverse range of skin tones and 2) replace names within books with multicultural names taken from researched databases to represent people of color and include greater representation in books. When implemented globally, we will introduce diversity to children around the world at a young age and help minorities become accurately represented in common books and literature.

Our education system is equitably flawed. When reading materials have illustrations of predominantly White characters and all names in novels and textbooks are common White names, children of color develop damaging implicit biases that only increase throughout their thirteen years of school education. With 75% of main characters in books being white (https://www.childrensliteratur...), children of color across the US are victims of these internalized implicit beliefs that they are marginal; these biases translate into flawed actions in society later on. Escalating systemic racial tensions are rooted in these harmful curriculums publicly backed by schools.

Students start school at the age of five when the brain reaches peak development, forming their perception of the world when susceptibility to detrimental biases is maximized. Our solution, Equify, tackles the implicit bias that most young students subconsciously develop through their schooling. The natural inclination to associate privileged races with good deeds and importance stems from the backbone of learning: novels and textbooks. As children continue through education and into their careers, their biases only feed into the perpetuating cycle of misrepresentation. Fictional characters in reading material should be representative of our globalized world, instead of only reflecting a portion of students.

Equitable education is the linchpin to working towards a more equitable, welcoming society for all. Thus, we present Equify, an application software that will use machine learning to replace common White names in educational reading material and alter skin tones in children’s picture books, with the purpose of eliminating implicit bias in children’s minds and making learning more equitable. Our solution will take data from credited databases on diverse names and randomize skin tones within existing children’s reading material. Equify’s algorithm will use text processing technology to parse through the words in a book to identify and replace commonly used White names with culturally diverse names. At the same time, it will identify the specific characters in the book through text processing and keep their names consistent throughout the whole book to eliminate any confusion to the user. Similarly, we will have a separate algorithm that will use image recognition to detect all characters’ skin tones and replace the skin colors with randomized skin shades to promote diversity in children's storybooks. This algorithm will identify specific characteristics of a person to identify the character and maintain that consistent skin color throughout the rest of the book.

Equify ultimately serves the American society, specifically catering to American children of color, ages 3-8 years. Children’s books and textbooks used in core school curriculum are largely catered to the White population of America with implicit and explicit bias in the lack of representation of non-White, non-heterosexual, non-male characters. This introduces bias among children without them even realizing it, often leading to children associating “white”, "male", and "heterosexual" attributes with “success”. By using image processing to change the skin tones and text processing technology to replace common White names with diverse names, Equify diversifies educational reading materials at the digital level for a lower cost, bringing marginalized children to the front of stories along with commonly represented White children. This is essential in reaffirming to children from underrepresented backgrounds that their stories matter. Equify’s impact extends to reducing systemic racism by targeting the most foundational source of these biases’ development: education.

Systemic racism stems from detrimental human biases, which are upheld by the lack of representation in educational reading material. Equify works to actively minimize these human biases, particularly in educational settings. By utilizing ML to modify character names and skin tone in digital storybooks, math textbooks, ebooks, and other reading material, Equify diversifies the current education curriculums and ensures that American children, our target population, are introduced to many different cultures and skin tones at an early age, leveling the playing field and diminishing the idea that standard White names and lighter skin tones are superior.

Equify is currently in the concept stage of development. We have defined the problem and are now in the research and exploration stages. Our team will begin development by implementing existing ML algorithms this summer with the help of a mentor who has 15+ years of experience in the ML field. We are on track to have a prototype ready by August 2021.

Our solution tackles the issue of race and implicit bias at the starting point: childhood. By changing reading materials of the younger population to include more diverse characters, we are introducing children to the representation of all cultures and skin colors at an early age. This will prevent biases from forming at a young age. 

Furthermore, it is an efficient and cost-effective way to change the current systems in place. Going through and manually changing the skin tones and names of characters for multiple children's storybooks will take up too much valuable time and money. Our solution utilizes machine learning to shorten the process by more than half the time and cost. By automizing the process of increasing representation in the educational curriculum, we are able to influence a generational change that begins early in the perception of race relations, stereotypes, and systemic racism.

1. Currently, we are serving 0 people since we are only in the concept phase.
2. In one year we aim to serve 1800 students, or one whole school.
3. In five years we aim to serve 30,000 students, or one whole school district.

We will measure our progress towards the goal by keeping track of how many books we have parsed through with our software, Equify. We will also be keeping a close look at the scores of Harvard's Implicit Association Test to identify trends during the research study we plan to conduct.

5 students working the summer to prototype our idea and 1 mentor guiding us through the process.

As recent high school graduates and oldest siblings, we have not only recently experienced the lack of representation in reading materials, but have also seen the impact it has on our younger siblings. Additionally, this past year, all of us students took a course on Margins in Literature and spent weeks on researching children's books and the influence the lack of representation has on young children. This led us to be well versed on the issue. 

We all also have experience working on computer science projects together. We have competed in four hackathons + competitions together through which we have built websites and an app which used AI. We will be partnering with a mentor who has 15+ years of industry and AI experience for this project and he will guide us through implementation of our first prototype. With our experiences and our mentor's help, we believe we are well-positioned to build this solution as well as introduce it to schools and ebook providers globally.

Our core leadership team includes five high school students of color. Four of us are women in tech. We have also consulted our school's equity board on our idea and refined it using their feedback. Our equity board includes teachers who identify as African American, Filipino, Indian, and many other ethnicities/races. 

Equify currently is a dream for our team. It's a concept that we know could have a huge impact for children like us, but one that we don't yet have the resources to build. With the network and mentorship that Solve offers, we hope that we can achieve our mission of building and releasing the Equify program, and can start helping learning communities integrate more diversity immediately. Solve will not only provide us with invaluable mentors that can help us create our program, but also with a network that can help upscale Equify and bring our solution to schools across the US.

As incoming college freshmen with no professional entrepreneurial experience, our team would benefit greatly from having Business Model and Public Relations mentors that can assist us in developing a brand for Equify and positioning our product to the schools and students that will benefit most from our solution.

In our minds, Equify can change individual lives, but we don't have clear evidence to prove its impact. The resources to conduct research and measure the impacts of Equify's solution are essential to determining the effectiveness of Equify, and are important milestones that we hope to reach in Equify. Mentors in Measuring Impact and Product Distribution can help us effectively determine the impact of our solution and use those results to pitch our project on a wider scale.

In terms of the Technology, we are sure that the text processing technology and image recognition machine learning models are achievable, as we've seen in Snapchat filters that change skin tones and Grammarly writing software extension that helps in writing style. With that knowledge, Solve Technology mentors to help us adapt these existing algorithms to build Equify would be invaluable to our product.

Our team behind our software solution, Equify, is qualified for this prize because of the ingenuity in our team, the impending racial crisis, and the tangible solution that Equify brings. Equify rewrites the reading material used in public education to reflect the diversity of America’s students with machine learning. The racial crisis we face today is the culmination of a generational lack of representation within our education system, which further perpetuates damaging stereotypes and perceptions. To solve the root of implicit biases that infiltrate our globalized society's dynamic, we need to reform our education system to make it more inclusive and equitable. Such biases can form as early as elementary school, a tender age where students are just beginning to read and consume media. However, the books and textbooks that children are exposed to are inherently misrepresentative of the globalized world we live in, where associations like lighter-skinned individuals being "good" are still a key characteristic of the majority of such stories. Equify takes such stories (in e-book format) and alters the names to be more culturally diverse and the complexions to reflect a broader range of skin tones in order to overcome implicit associations students form at early ages. Our team is dedicated to build our solution as soon as possible to impact as many generations of students as possible, preventing this systemic cycle of racial violence from repeating. As a team of incoming undergrads, we need all the financial support we can receive to build our solution robustly.

Our team behind our software solution, Equify, is qualified for this prize because of the ingenuity in our team, the impending racial crisis, and the tangible solution that Equify brings. Equify rewrites the reading material used in public education to reflect the diversity of America’s students with machine learning. The racial crisis we face today is the culmination of a generational lack of representation within our education system, which further perpetuates damaging stereotypes and perceptions. To solve the root of implicit biases that infiltrate our globalized society's dynamic, we need to reform our education system to make it more inclusive and equitable. Such biases can form as early as elementary school, a tender age where students are just beginning to read and consume media. However, the books and textbooks that children are exposed to are inherently misrepresentative of the globalized world we live in, where associations like lighter-skinned individuals being "good" are still a key characteristic of the majority of such stories. Equify takes such stories (in e-book format) and alters the names to be more culturally diverse and the complexions to reflect a broader range of skin tones in order to overcome implicit associations students form at early ages. Our team is dedicated to build our solution as soon as possible to impact as many generations of students as possible, preventing this systemic cycle of racial violence from repeating. As a team of incoming undergrads, we need all the financial support we can receive to build our solution robustly.

Our team behind our software solution, Equify, is qualified for this prize because of the ingenuity in our team, the impending racial crisis, and the tangible solution that Equify brings. Equify rewrites the reading material used in public education to reflect the diversity of America’s students with machine learning. The racial crisis we face today is the culmination of a generational lack of representation within our education system, which further perpetuates damaging stereotypes and perceptions. To solve the root of implicit biases that infiltrate our globalized society's dynamic, we need to reform our education system to make it more inclusive and equitable. Such biases can form as early as elementary school, a tender age where students are just beginning to read and consume media. However, the books and textbooks that children are exposed to are inherently misrepresentative of the globalized world we live in, where associations like lighter-skinned individuals being "good" are still a key characteristic of the majority of such stories. Equify takes such stories (in e-book format) and alters the names to be more culturally diverse and the complexions to reflect a broader range of skin tones in order to overcome implicit associations students form at early ages. Our team is dedicated to build our solution as soon as possible to impact as many generations of students as possible, preventing this systemic cycle of racial violence from repeating. As a team of incoming undergrads, we need all the financial support we can receive to build our solution robustly.

Our team behind our software solution, Equify, is qualified for this prize because of the ingenuity in our team, the impending racial crisis, and the tangible solution that Equify brings. Equify rewrites the reading material used in public education to reflect the diversity of America’s students with machine learning. The racial crisis we face today is the culmination of a generational lack of representation within our education system, which further perpetuates damaging stereotypes and perceptions. To solve the root of implicit biases that infiltrate our globalized society's dynamic, we need to reform our education system to make it more inclusive and equitable. Such biases can form as early as elementary school, a tender age where students are just beginning to read and consume media. However, the books and textbooks that children are exposed to are inherently misrepresentative of the globalized world we live in, where associations like lighter-skinned individuals being "good" are still a key characteristic of the majority of such stories. Equify takes such stories (in e-book format) and alters the names to be more culturally diverse and the complexions to reflect a broader range of skin tones in order to overcome implicit associations students form at early ages. Our team is dedicated to build our solution as soon as possible to impact as many generations of students as possible, preventing this systemic cycle of racial violence from repeating. As a team of incoming undergrads, we need all the financial support we can receive to build our solution robustly.

Our team behind our software solution, Equify, is qualified for this prize because of the ingenuity in our team, the impending racial crisis, and the tangible solution that Equify brings. Equify rewrites the reading material used in public education to reflect the diversity of America’s students with machine learning. The racial crisis we face today is the culmination of a generational lack of representation within our education system, which further perpetuates damaging stereotypes and perceptions. To solve the root of implicit biases that infiltrate our globalized society's dynamic, we need to reform our education system to make it more inclusive and equitable. Such biases can form as early as elementary school, a tender age where students are just beginning to read and consume media. However, the books and textbooks that children are exposed to are inherently misrepresentative of the globalized world we live in, where associations like lighter-skinned individuals being "good" are still a key characteristic of the majority of such stories. Equify takes such stories (in e-book format) and alters the names to be more culturally diverse and the complexions to reflect a broader range of skin tones in order to overcome implicit associations students form at early ages. Our team is dedicated to build our solution as soon as possible to impact as many generations of students as possible, preventing this systemic cycle of racial violence from repeating. As a team of incoming undergrads, we need all the financial support we can receive to build our solution robustly.