Month: March 2016

Sentiment Analysis in Python with TextBlob

Welcome to the eleventh blog of 52 Technologies in 2016 blog series. If you are following this series then you would have probably noticed that I already wrote week 11 blog on tweet deduplication. I was not happy with the content so I decide to write another blog for week 11.

In week 11, I decided to spend time to learn about text processing using the Python programming language. We will only focus on Sentiment Analysis in this blog. I have written about sentiment analysis multiple times in last few years. We learnt how to do sentiment analysis in Scala using Stanford CoreNLP in week 3 blog. Sentiment analysis gives you the power to mine emotions in text. This can help you build awesome applications that understand human behavior. Few years back, I built an application that helped me decide if I should watch a movie or not by doing sentiment analysis on social media data for a movie. There are many possible applications of Sentiment analysis like understanding customer sentiment for a product by analysis of reviews.

You can read full blog here https://github.com/shekhargulati/52-technologies-in-2016/blob/master/11-textblob/README.md

Gatling: The Ultimate Load Testing Tools for Programmers

Welcome to the tenth blog of 52 Technologies in 2016 blog series. Gatling is a high performance open source load testing tool built on top of Scala, Netty, and Akka. It is a next generation, modern load testing tools very different from existing tools like Apache JMeter. Load testing is conducted to understand behavior of an application under load. You put load on the application by simulating users and measure its response time to understand how application will behave under both normal and anticipated peak load conditions.

Gatling can be used to load test your HTTP server. HTTP is not the only protocol that one can load test with Gatling. Gatling also has inbuilt support for Web Socket and JMS protocols. You can extend Gatling to support your protocol of choice.

Load testing is often neglected by most software teams resulting in poor understanding of their application performance characteristics. These days most software teams take unit testing and functional testing seriously but still they ignore load testing. They write unit tests, integration tests, and functional tests and integrate them in their software build. I think part of the reason developer still don’t write load tests has to do with the fact that most load testing tools are GUI based so you can’t code your load tests. They allow you to export your load test as XML.

You can read full blog at https://github.com/shekhargulati/52-technologies-in-2016/blob/master/10-gatling/README.md