The trail recommendor aims to provide optimal recommendations for public trails in the region Washington for a particular user.
For example,
"I am an enthusiastic hiker but a busy porofessional in Seattle city who wants to explore nature on weekends but avoid the crowds".
This recommendor will provide me the best trails available in the area according to my preferences.
In order to get recommendations using this project, follow the guidelines below;
- Install
streamliton your machine - Run
streamlit run demo.py - Once the window pops up, enter your zipcode, distance to drive, date and time of your visit.
The goal of this project was to collect historical weather, trails information, and hikers' foot traffic on the trails.
We used the following to accomplish that;
| Source | Description | Type | Key metrics |
|---|---|---|---|
| Washington Trails Association | Washington Trails Association is a non-profit organization that advocates protection of hiking trails and wilderness, conducts trail maintenance, and promotes hiking in Washington state. | Web scrapping | Elevation gain, Distance, Difficulty level, Location etc. |
| Trailforks | Trailforks is a trail database, map & management system for users, builders and associations. A platform for trail associations to keep track of trail conditions, builders to log work & users to discover, plan and share their activities. | Web scrapping | User checkins for year, month, date, and hour of a day. |
| Visual Crossing | Visual Crossing is a leading provider of weather data and enterprise analysis tools to data scientists, business analysts, professionals, and academics. | API | For a particular day and hour; Sunny, Rainy etc. |
The project uses knn clustering to arrange trails with similar usage and popularity and predicts the ones with lower popularity within hiker's given radius.