Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

A system for prediction of epidemic spread of airborne diseases using location data from social media feeds

IP.com Disclosure Number: IPCOM000246332D
Publication Date: 2016-May-31

Publishing Venue

The IP.com Prior Art Database

Abstract

In this article, we propose a system that uses real time streaming, data analytics and spatio-temporal technology to determine and predict the spread of the airborne diseases by making use of the locations of the social media posts.A social media post is the activity that is done by an user on a social media platform such as facebook, twitter, instagram etc. The system can correlate patient data with others who were in the same place (or) who crossed paths with the patients using geospatial analytics in spatio-temporal dimensions, to find out the probable candidates who may be exposed to infection. Using this kind of system, a preemptive and preventive medical treatment could be initiated to arrest the spread and address any such possible outbreak.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 36% of the total text.

Page 01 of 17

A system for prediction of epidemic spread of airborne diseases using location data from social media feeds

This article proposes a system to determine and predict the spread of the airborne diseases by making use of the locations of the social media posts. A social media post is the activity that is done by an user on a social media platform such as facebook, twitter, instagram etc. The location of the posts is then subjected to geospatial analytics and temporal analytics to arrive at decisions that can predict the spread of an airborne disease.

In this article, we propose a system that uses real time streaming, data analytics and spatio-temporal technology and data available from social networks. Here system can correlate patient data with others who were in the same place (or) who crossed paths as that of the patients using spatio-temporal dimensions, to find out the probable candidates who may be exposed to infection.

Using this kind of system, a preemptive and preventive medical treatment could be initiated to arrest and addressed any such possible outbreak.

A patient gets diagnosed in a hospital that he has been affected by a airborne disease.

I. Identifying location of the patient

The health authorities then get the social media ids of the person on his acceptance. This id is now used to filter out his activities on social media for past n number of days. n can be defined as the time duration where a disease organism will remain virulent and is capable to spread through air. From various social media feeds, the data is streamed and the activities by the patients are filtered. From this data, the geographic location in latitude and longitude of the patient is parsed and used. This would mean that we now have the locations of the patient activity which implies the patients actual location for the past n number of days in which the organism remained virulent.

II. Finding people who were in the proximity of the social media posts

This article proposes three ways of analysis.

A. On the Spot analysis


1. The following image shows the posts from different users including the patient. The patient's post is shown in red in the below figure.

1


Page 02 of 17


2. For every geographic location from where the social media posts were made by the patient, a circular area of coverage of X meters is determined. X is the maximal distance from a patient where a person can acquire the disease as it is airborne. In the figure below, there are two posts in green denoting that the users who made their posts are not at risk because they were not within the distance X. However, there are posts in blue that are within the area of infection and could be possible candidates who may have been exposed. This is the geo-spatial analysis. Here you see three posts in blue that denotes probable candidates who got exposed.


3. However just because some posts are within X meter distance, it may not mean that they should definitely be exposed as they could have a...