ISSS608 2017-18 T1 Assign RACHEL TONG Preface

Home

Preface

Origins

Transmission

Containment

Coda

Preface - Analysis Strategy

Based on the data provided, we establish our knowledge of:

1. The existence of a flu-like disease
2. An approximate spatial/temporal understanding of various events at a person-level view

As such, the strategy for this investigation about the mysterious outbreak will be centered on the following aspects.

Blind Hypotheses about Disease

Possible Disease Origins

Contextually, there is a low probability of local-based animal source due to lack of agricultural dependency by economy. More probable disease origins could include human sources or exposure to biologically hazardous substances. Some specific scenarios (blindly hypothesized) are as follows.

• Human sources
  • Isolated cases from travels abroad
  • Bad weather could have caused a few source-individuals to fall ill
• Exposure to Biohazards
  • Plant/Factory Explosion
  • Food Poisoning
  • Water supply contamination via sewage/river
  • Biohazard spillage or leaks in public areas

Disease Transmission

For communicable diseases, transmission could be by either human-to-human contact, by air, or by water. It is also possible that some symptoms may be purely an effect of biohazard exposure and may thus be non-communicable.

Human-to-human

In context of communicable diseases, human-to-human transmission would possibly be identified in the presence of notable symptom-based patterns with congregations of people. Relevant areas of interest would include Vastopolis Dome and Westside Stadium, the Convention Center, the State Courthouse, and the Capital Building. It should also be noted that the central area, comprising mainly the Uptown and Downtown regions, have the highest percentage of daytime human traffic influx (290% and 190% of their baseline populations respectively – see graph below). The next region with a substantial increase in daytime population is Westside (90% increase).

<insert image>

Considering the daytime population (thin blue bars) with respect to the percentage increments over the baseline population densities of each region, it may be seen that Westside and Downtown have the second and third largest daytime populations respectively. Moreover, despite not having data about the exact size of each region, the Downtown and Uptown areas (at map centre in yellow outline) are visibly smaller than most other regions; comparable to only the Smogtown and regions Cornertown (bottom and top left corners, outlined in light orange). In contrast, the daytime population of Smogtown and Cornertown are still much smaller compared to the Uptown and Downtown regions.

<insert image>

Additional points to note are that besides Cornertown, Villa, Riverside, and Smogtown, the rest of the regions have a sizeable efflux of human traffic of more than 10% in the daytime as seen by the negative percentage change in daytime population. It is highly likely that the central region is the destination of these daytime commuters, given the presence of the important administrative buildings (Capital Building, State Courthouse, and Convention Centre) in the central region, and the presence of the Westside Stadium in Westside. Given these high-concentration regions, symptom-related patterns originating in these locations could indicate the existence of human-to-human transmission.

Air-borne Transmission
Considering any correlations between wind direction and symptom-related messages would indicate whether the disease is likely airborne.

Water-borne Transmission
Given the fact that one of the city’s water sources is the south-flowing river, and that major water-activities are also carried out at the river, a major water-borne disease would likely originate either from along the river or from reservoirs. It would also hit those living by the banks of the river with greater concentration and earlier than it would the rest of the city on the assumption that those living near the river are more likely to engage in water-based activities.

The Facts

Through personal sentiment derived from geo-tagged microblog data, the microblog file data allows phrase-based mentions to be tracked geospatially over time in order to identify patterns in the data. In particular, we are interested in patterns involving Symptoms, and patterns involving major events within Smartpolis(especially those involving congregations of people, which are likely to give insight to human-to-human type disease spread hypotheses).