In this article, we will explore an implementation on the Hierarchical Factorization (HPF) model in Edward. This implementation will use the movilens 2k dataset to learn latent features that try to capture the preferences and attributes of users and movies respectively. The code is avalaible on github.
Web data mining is a major component in the data science activities landscape. With its use, we can track brand mentions from social media statuses, policy sentiment on forums and "hot" topics from political blog posts.
In this article, we'll ding into the posts from the politics's rubric of the site Journal Du Cameroun (JDC) to find what are the topics developed there. The scraping was done on May the 10th 2017 and the dataset includes 136 articles between February 17, 2017 and May 9, 2017.
In this post, we introduce Mesa a library written in Python3 for Agent-Based Modeling (ABM). Agent-based modeling is a growing topic of interest in many scientific disciplines, due to its flexibility for modeling complex dynamical systems. Agent-based models (ABMs) consist of autnomous, interacting computational objets, called agents within a given environment.
Previously, we explore topic modeling an algorithm used to discover what users are talking about. But often, its not important to know what users are saying, but how they are saying it. Sentiment analysis seeks to automatically associate a piece of text with a sentiment score, a positive or negative emotional score. Aggregating sentiment can give an idea of how people are responding to an event or a topic.
In the social network analysis part, we explored a model that exploits the links between the entities to help us find the key players in the data. Here, we will focus on the tweet’s text to better understand what the users are talking about. We move away from the network model we’ve used previously and discuss other methods for text analysis. We first explore topic modeling, an approach that finds natural topics within the text. We then move on to sentiment analysis, the practice of associating a document with a sentiment score