FACELESS

Inspiration

FACELESS is a non-traditional pair-making(dating) website. We think people nowadays are paying too much attention to the appearance instead of what's inside of a person. Thus, we are here to get rid of the fancy photos, no more “5 Professional Steps to Taking That Perfect SELFIE!” and help people to explore other's inner beauty, as well as themselves.

What it does

Without actual photos, people will create comic faces of themselves according to their descriptions. This way of abstracting the appearances can force users to pay more attention to other features of people. Meanwhile, face making can be a lot of fun as well! We will provide many different features, so it will have details to some extends(eye color, hair style, eyebrow, skin tone...). One annoying thing we often find in traditional pair-making websites is that, sometimes users are recommended to people who they are completely not interested in… (We all have high expectations!) Our system will provide a double-sided selection, which will ensure that you are only recommended to people that you will be interested in. By analyzing what kind of invitations you have rejected, we can see what kind of people you are not interested, and there will be an implicit avoid type calculated for you. We will also track future activities by questionnaires to modify our recommendation. So no more “junk invitation”!

How we built it

we use vectors to represent persons. By using the idea of projection, dot product and inner angle, we calculate exactly how far two vectors are separated, in other words, how different two people are.
Each user has three vector associated with them. One represents the user. This is based on the input data of the user as well as the feedback from other users who they interacted with. One represents what kind of people the user is looking for; this is also based on the input data of the user and will be constantly adjusted as the user building connections with other people. The last one represents what kind of people the user does not like; this is calculated by collecting the data of persons the user has rejected. One of the most important feature of our design is that each operation you make will update your own unique data. Why is this important? Because sometimes, people do not know themselves that well! While describing their anticipations, people tend to idealize some features, and sometime these are not what they really want. By looking into what kind of invitation one accepted, we will adjust the “prefer” vector; by getting the feedback from other users, we will adjust the “self” vector; by collect rejected invitations, we will adjust the “avoid” vector accordingly.

Challenges we ran into

website design is kind of new for us, and figuring out the adjust algorithm is difficult.

Accomplishments that we're proud of

We have had a good starting point on this project, and now we have a much clearer idea of what to do next. As web developing beginners, We also successfully built a webpage with links from scratch.