Pročitajviše
Pročitajviše
Pročitajviše
arrows
arrows
arrows
Pogledajvideo
Posjetiblog
Posjetiusluge
Posjetinaš rad
Back to Work
dots
Upravljanje velikim podacima

NSoft: Using Machine Learning to Discover Competition

Read more about machine learning project, where we made competition and potential partner analysis for premium sports betting software provider NSoft.

arrow down Scroll down

Company Background

NSoft is an award-winning software solutions provider for the gaming and betting industry. The company has more than 200 employees, 100+ companies as its customers, and is currently operating in 35 countries.

Expectations

The client wanted to make a worldwide market analysis to find new potential partners and detect competition worldwide.

Analysis results would serve as a really good input for the sales team since it would give them an overview of market size and expansion opportunities.

Additionally, competition analysis might discover markets that the company is not yet present.

Goals

Find new market & partner expansion opportunities.

Discover competition and analyze the services they provide.

Solution

The initial part of the project was to research Natural Language Processing (NLP) and text classification algorithms. After successful research, the proof of concept was built and demonstrated to the client. 

Machine learning was used to find new potential partners and competitors. We fed the program with data so it can start to classify websites into two separate groups – partners and competitors.

Web crawling was done on over 230 million websites in a period of one month with a classification accuracy of 92%. To avoid certain location restrictions, we used residential proxy servers.

Results

230M
Websites crawled
92%
Classification accuracy

Newsletter

Ako ti treba više informacija, bez brige. Izradili smo pregledni dokument koji objašnjava proces i detalje ove studije slučaja. Ostavi nam svoju email adresu i odmah ćemo ti poslati kopiju.