WEB SCRAPING SERVICE

Initially, the client – a large investment company – requested review collection to a particular brand name. After the analysis of web resources containing reviews to this company, we proposed a plan: 1 keyword (brand name), 9 web review platforms, 1 report per week, payment required 1 time per month. This case led us to the idea of a paywall based review monitoring tool.

The customer needed to extract information on transport accidents in a number of countries from Wikipedia pages. In his insurance app, he would then use this statistics to evaluate the risk of travelling to different regions of the world. The unusual on this case was that we had to process data in .json format.

The client reached out to us having a clear idea and clear technical requirements: a data analysis company from South East Asia wanted to scrape job information from Japanese job portals. The requirements allowed us to choose one of three job portals depending on their scrape-ability. Figuring out the Japanese hieroglyphs and processing a giant data set of 400000+ units were main challenges in this case.