Data Story
a. Data & analysis/method
A simple idea of how many crocodiles seen and recorded in 7 months to the ratio of number of crocs estimated to be in the NT, using the total number of crocodile sightings (in the CSV file, Crocodile Monitoring Survey 2021-2022) divided by the total number of crocodiles estimated to be in the NT
100000 estimated crocodiles In the NT (NT Gov BeCrocWise, N.D) 5282 spotted (in dataset) in a range from 21/06/2021 –> 25/08/2022 (7 months, some months not included) that’s 19% of the estimated population of crocs in the NT (100000/5282 = 18.932%)
That tells us that somewhat 19% of the crocodile population out of 100000 was seen in a span of 7 months according to data, that also tells us that there are many more out there that has not been seen and recorded. This can also showcase how the crocodile's intelligence and sly nature plus the murky waters of the top end makes it harder for human and IT observation to spot and record crocodiles.
b. Possibility of crocodile presence:
Our idea is using machine learning to predict the possibility of seeing crocodile in a specific area, based on historical data, this function will be shown in the app.
If the possibility runs from 1%-100%, which mean that area may have crocodiles, the app will show the red zone.
If the possibility is 0%, which mean that area is “crocodile-free”, the app will show the green zone.
The more precise and timely data we have, the more accurate the app will be in crocodile prediction of presence.
In the future we are aiming to provide Realtime data-analysis, for example, if crocodile present or enter in a green zone, the app will send an alert to people in that green zone and change the zone into red.
c. Visualize crocodile presence
We use NodeJS and Google map API to visualize datapoints from Excel file (Crocodile Monitoring Survey 2021-2022)
The map utilized the ‘longitude’ & ‘latitude’ datapoints (5282 datapoints) recorded in the file given, each dot represents a crocodile sighting that has been recorded. (Details on Report)
d. Future data collection:
The current data is minuscule and data sittings of crocs are limited as data used has only been recorded from 2021-2022 and not the current date. So, in essence our idea is that we need to collect data by Technology (drones/ security cameras), human reports from the app solution that will be provided, this can most likely be done with the help of big organisations, government and Parks and Wildlife Commissions of the NT, lastly by crowdsourcing.