Online Sales Research for Energy Safe Victoria: A Pilot Study

The project aims to researching into the type, efficient and cost effective of an image-based algorithm. The pilot study interrogated textual information posted by sellers, who may be parallel importers. Combining this with the images from their advertisement will give ESV a much richer dataset to identify the products for matching safety purpose. This will help ESV to understand:
•limitations on the imagery on the EESS database (which may be used to guide future database requirements)
•differences between sellers as to textual and other information available through their APIs
•differences between sellers in the volume and quality of information available on products
•ability to get AI running on a single product and then test the methodology by broadening to a wider product range.