V camp

A Full Stack web app with a recommender system to view, add, modify, review, rate, locate campgrounds all over the world.

There is zillions of pieces of data around us which can be effectively used to train computers to perform human-like tasks. A vast majority of human understandable data is textual in nature. A Machine learning model can be trained to build Recommender systems which can recommend things based on the similarity between the textual description of things. V-camp uses a Machine learning model which recommends camps based on the location and textual description of camps and similarity with other campgrounds.

It uses concepts such as TF-IDF Vectorization, Tokenization, Cosine similarity to convert textual data into machine readable format and train a machine learning model off of it.

You can view the project here -> V camp

• Click on View Campgrounds button
• Click on view button for any campground (alternately you can locate campgrounds on the map)
• Scroll down to Similar places to visit.. section

Data Collection

The data was scraped from a travel website using BeatifulSoup and requests library. The data attributes include name of place, textual description, location, rating, and image url. The data was converted into a pandas dataframe for convenient handling of data.

Recommender System

• The recommender system takes in a campground and filters other campgrounds by same location.
• The filtered campgrounds data is then passed to a tf-idf vectorizer to convert textual description to numeric vectors.
• The final step is to use cosine similarity to find campgrounds within same location using the campground descriptions.
• The output is five most similar campgrounds.