Project 2 -Tweets of Conjunctivitis

Data set : Twitter data collected for Conjuctivitis.

My narrative is author based. Social media can be used as a disease surveillance tool. By looking at the number of people tweeting about conjunctivitis, can indicate rise in the number of cases of the disease. My main motivation behind the visualization is to use social media data to identify the disease pattern of occurrence over a course of the year, and compare it with ground truth i.e. information from public health agencies.

People tweet their lives. Social media platforms promise increased opportunities for a more timely and accurate spreading of information. Time series pattern of of tweets collected between 2012 to 2014, the peak of the diseases occurs during the early spring. This pattern matches with the hospital records. Indicating the rise in conjunctivitis during the early spring.

These new sources of analysis and information are intended to complement traditional sources of epidemic intelligence. Further social media and web searches help identify individuals within a network who should be targeted for vaccinations to prevent the spread of disease. GIS analysis can, for example, locate a potential disease hot spot and calculate a nearby hospital’s ability to handle the expected increase in service demand if an outbreak should occur

To improve on this visualization, I would like to add a tooltip to see the exact numbers, and brush to select the range of the dates to visualize. Also Geo tags of tweets could be used to plot them in a map. This helps isolate the network of individuals, and a potential outbreak.

Dealing with all the noise and picking the correct signal is a challenge. Since people just tweet very generically, and do not write specific symptoms and causes. Maybe it’s an initial indicator of the disease outbreak and health officials can take preventive measures to avoid that region of the epidemic.