Reddit.com is a social media news and discussion forum, interesting data can be pulled from Reddit that can show real-time trends of popular topics.

-There are thousands of users on the platform, commenting on different topics and forum, expressing their views. As a mass media platform, it can be weaponized into a powerful tool that can manipulate the masses.

-As part of our project, we will be classifying accounts as either normal users or bots on Reddit.

-Our data source is Reddit API. We are planning to utilize RedditExtractoR “https://cran.r-project.org/web/packages/RedditExtractoR/”.

-This API returns contents of all of Reddit ecosystem, subreddits, user data and Karma, upvotes and downvotes, trending posts.

-We intend to use the API data in JSON format. We plan to analyze a few subreddits and the comments posted on them, to classify the types of users.

-For cleaning the data, we will be making use of tidyverse package to remove outliers and ensure data is tidy before analysis.

-The text of each comment will be referred to as a document and each comment will also have a corresponding label, i.e. normal or bot. we intend to convert each comment into a bag of words and represent each comment as a vector.

-We plan to make use of different classification models to predict whether each comment is a bot, a troll, or a normal user

We plan to compare models above and find the best model among them. To evaluate the quality of the output, we use RMSE, confusion matrix, prediction, recall and F scores.