# Context This is a collection of data scraped from MountainProject.com and Trailspace.com. # Content Details about how the data was scraped can be found on my GitHub: https://github.com/pdegner/DL_final_project There are 5 datasets here: Routes Route: The name of the route Location: The location and sub-location of the route URL: Link to the Mountain Project page about this route Avg Stars: The average rating of a route Route Type: Only sport and trad routes were scraped Rating: How difficult the route is Pitches: How many pitches the route is Length: How tall/long the route is Area Latitude/Longitude: GPS coordinates of the route desc: Description of the route protection: What gear is needed to climb the route num_votes: How many people have voted on the quality of a route Trailspace brand: Brand of item being described model: Model of item being described rating: The rating that the reviewer gave rating_text: The description of the item written by the reviewer Discussion and Review: These are forums on Mountain Project topic: The main topic of a posting page\_num: What page of the topic the comment was made on post\_num: Where on the page is the post text: Text of the post join\_date: When the poster joined Mountain Project post\_date: When the comment was made num_likes: How many likes the comment received Labeled For a school project, I created a model that analyzed the sentiment of the forums. Details about that model can be found in the GitHub REPO linked above. Overall, it achieved 81.6% accuracy in labeling the forums as positive (2), negative (0), or neutral (1). topic: The main topic of a posting page\_num: What page of the topic the comment was made on post\_num: Where on the page is the post text: Text of the post join\_date: When the poster joined Mountain Project post\_date: When the comment was made num\_likes: How many likes the comment received pred\_label: The label my model predicted for a posting true\_label: I manually labeled ~4000 examples to train and test the model # Acknowledgements Big thanks to MountainProject.com for letting me use this information. # Analysis Ideas Do climbers that have been members of Mountain Project for a long time prefer a different type of gear than those that are newer? Has the sentiment of climbing brands or gear changed over time? Are high-quality routes given harder ratings?