Hello! My name is Blaz, and I am an aspiring developer with a passion for Python and Java.
Here are a few of the projects I have worked on:
Trackify is a web application that allows users to track prices of items and products across multiple online retailers (mostly focused on Slovenian retailers, although Amazon.de is supported too). This project was built using Python, Django, and JavaScript, and it utilizes a PostgreSQL database to store data. The repository also includes a Docker setup for deployment. Some of the features of Trackify include:
- Simple and friendly UI
- Graphical representation of price movements
- Registration and saving of items into user profiles
In this project, I explored the use of Naive Bayes models to predict the sentiment of news headlines. I used the Manually Sentiment Annotated Slovenian News Corpus SentiNews 1.0 to train and test my models. Some of the techniques I used include:
- TF-IDF
- Pandas library for corpus analysis and cleanup
- Seaborn library for graphical representation of dataset and model accuracy (confusion matrix)
- Scikit-learn and their Multinomial Naive Bayes and Bernoulli Naive Bayes classifiers
A prototype of an open forum/discussion board, built with Django. It has the following features:
- Session/cookie based voting system
- Rich text editor for comments/posts
- Pagination
- Search functionality