Carpe Diem Ski Lab

Inspiration

Finding new ski friends is hard - some people you vibe with and others you just don't. For adult beginners who don't know anyone else who skis, that's even worse! I created this project because I used to be a completely new snowboarder who didn't know anyone else to shred snow with, and even though I'm an extrovert who doesn't mind meeting new people on trips, I wanted a way to create stronger, more consistent connections with friends who would help me grow as a person and shared my purpose and motivations for enjoying the mountain life.

What it does

Answer a few thoughtful questions, and we'll use data science to introduce you to new local ski-friends who share your specific values and motivations for skiing!

How we built it

Given the 24 hour time constraint of the hackathon, we decided to focus on declaring 3 popular reasons or value drivers, for why people would go skiing/snowboarding: for fitness, personal growth, or interpersonal connections. Many utilize all 3 as values, however for simplicity we focused on separating out into 3 groups - each group contains all of the users that hold that item as a value. These groups are therefore not mutually exclusive.

We used data science to grab a dataset from Kaggle using OkCupid data and filter for common mutual information scores, or sets of two words that are most likely to occur together, for each set of users that have a given value.

We used this to find patterns in the bigger picture for common vocabulary or speech patterns per value.

We then used the differences in these patterns across the 3 values, fitness, personal growth and interpersonal, in order to come up with some better questions for Ski Lab users to answer, than those of the original dataset, that still achieve the mission of getting to know each person on a level deeper than the surface. We used design thinking and focus group principles in order to extract purpose behind the questions to tease out the larger pattern differences in vocabulary and speech patterns across the 3 value groups.

We then created a simple data collection survey via Typeform to ask our improved questions, and created a landing page for people to learn more about our mission of connecting people with new ski buddies who share similar motivations/values for getting out on the slopes.

People are now able to fill out this Typeform to answer questions that were built with our data science-informed principles in mind. Currently we will be performing simple searches manually to cross-reference their essays answering our questions, in order to determine which value of the 3, that they value the most. Based on this, we can manually email users with their curated introductions to friends who share the same value motivation behind going on ski trips.

Challenges we ran into

Not having enough time to build out a full model or automate the process

Accomplishments that we're proud of

Creating an accessible method for others new to the ski life to jump in to the cold as hell snow + camaraderie of the ski life

What we learned

Those who value interpersonal relationships highly, tend to speak often about other people, philosophy and life, and often focus on ways to see the "big picture" in life.

What's next for Carpe Diem Ski Lab

In the future, given more time we will be able to build out a robust model and automate sending out email or text introductions to friends who are recommended for one another.

Built With

• jupyternotebook
• kaggle
• python