ML & Natural Language Processing on 0.5 M Amazon Reviews
Our data set came in a sq-lite format which consists of 2 columns (review description and stars ranging from 1-5) with about half a million reviews about various products, I have used Python and NLTK to perform natural language processing on the review description to remove the stop-words and to perform stemming. Once the data set is natural language processed it is subjected to for feature extraction of processed text using CountVectorizer which is later passed onto a supervised classification algorithm(Gaussian Naive Bayes) to learn from the input labelled data-set for learning. We are not done yet, the data set is huge so you may have a question of how I handled it so here it is, Since countvectorizer does not support partial fit(we could use HashingVectorizer to solve but I need this to be done using CountVectorizer) and there are several other barriers which hinders to perform out-of-core learning ....CONTINUED SEE THE IMAGE. FREELANCER LIMIT FOR DESCRIPTION is 1000 CHARACTERS
About Me
Programming Skills: + Developing applications for desktop using C, C++, Java and Python + Building websites using Python, Flask, MySQL or MongoDB along with HTML, CSS and JS for frontend. + Building machine learning models like regression, classification and clustering to complex neural network models like convolutional neural networks and recurrent neural networks. + Building cross platform mobile application using React Native and supporting native mobile applications built with Java. + Building web-scrappers with Python and web-automation using selenium. + Developing restful APIs with Flask and graphql endpoints with Graphene. + Visualization with matplotlib, seaborn, plotly for web, etc., + Deploying Python, PHP, NodeJS, Java application on linux servers. + Working knowledge on Apache and Nginx. + Deploying and Building Docker images from scratch and debugging the running containers. + Proficient with Linux machines like CentOS, Ubuntu and Debian. + Working knowledge on cpanel and ispconfig. + AMQP with celery and RabbitMQ
