How I configured Authorized Access to Kibana Dashboards

Leave a comment

One most important missing feature of Kibana is authorized access to Kibana charts and dashboards, kibana/issues/1610. Implementing a workaround has been the most difficult part of my Twitter Analytics application. Check it out here. PROBLEM STATEMENT: I want to allow end-users to play around with the graphs but without affecting other users. Users must be… Continue reading How I configured Authorized Access to Kibana Dashboards

Analytics on #sjsu

Format Post format: ImageLeave a comment

Hi, here I am going to list out the analytics that I have derived from my application Twitter Analytics Using Elasticsearch and Kibana. I will describe the architecture behind this application in another post!

I have started collecting tweets with #sjsu on 28 January 2016. Till now, total number of tweets collected are 12,234. 

Note: In this post, when I say tweets, I mean tweets with a hashtag ‘sjsu’.

Case Study 1:
SJSU admissions rolled out on Febraury 15

On Febraury 15, 2016, there is a record with maximum number of tweets (534) collected per day! Some tweets are negative expressing why sjsu does not honor president’s day and others are all about the exhilaration for the admission decisions rolled out from SJSU. Also, a couple of people are unhappy for getting rejected.

People started tweeting about SJSU admissions at around 4:30 p.m. From the data collected, glenelazegui is the first person to tweet about the admission decision, congrats to him! Number of tweets peeked around 8:30 p.m. Total 164, i.e., nearly 31% of tweets are collected from 8:30 p.m. to 9:30 p.m.

Screen Shot 2016-03-16 at 1.38.50 PM
Number of tweets over time

 

Other hashtags that are most associated with #sjsu are #blessed, #golobos, #spartans

Most people used Twitter for Iphone, Twitter web client and Twitter for Android.

Screen Shot 2016-03-16 at 1.44.55 PM
Top Twitter applications

 

Overall, the positive and negative sentiment on this day was:

Sentiment analysis of tweets collected on Feb 15, 2016 (SJSU admissions)
Sentiment analysis of tweets

 

I am using TextBlob: Simplified Text Processing library to perform sentiment analysis. I think it is not doing a better job at classification as most of the tweets are positive.

A sample of positive tweets marked as neutral are:

Tweet sample showing happiness of students admitted into SJSU
Tweets sample showing happiness of students admitted into SJSU

 

Let me know in the comments, what other interesting analytics you can do with this data!