How legal is it to use data for machine learning obtained by scraping third-party websites?

Question

data mining

For example, I wrote a parser for the pages of some site with news, chess games, match results - it doesn't matter. Then, based on this data, I trained a neural network, and based on it I created a paid web service. How legal is this in terms of the copyright of the original data that I used?

Answer 1

Good afternoon, if your new service does not provide for the disclosure of any personal data, data related to trade secrets or violation of someone else's copyright, then it is quite possible. The collection of such data by scripts is quite equivalent to normal surfing, and their use is regulated by standard provisions (license, copyright, personal data). In the Russian Federation, no one has yet thought of licensing or prohibiting machine learning products that were obtained as a result of analyzing competitor data. But yesterday I heard that there is such a ban in Holland, but not in the USA. So, if you think about going to the int. market carefully analyze the legislation.