Оn the information resource when applying information technologies for information provision, the collection, systematization, and analysis of data relating to the preferences of Internet users located in the Russian federation are carried out.
This document contains the rules for the use of recommendation technologies on the website habr.com (hereinafter referred to as the “Site”).
The rightholder of the Site is Habr Blockchain Publishing Ltd. Registration number: HE 289873. Address: Diagorou 4, Kermia Building, 6th floor, flat / office 601, 1097, Nicosia, Cyprus.
Email address for legally significant communications: support@habr.com
1. Information Provided Using Recommendation Technologies
1.1. “Read Now” feed.
1.2. “Daily Habr” feed.
1.3. “Daily best” feed.
1.4. “Similar” feed.
1.5. Advertising banners.
2. Recommendations in the “Read Now” Feed
2.1. General description: a block containing links to the 5 (five) most viewed publications at the current time. A link to another publication may be added manually (without using recommendation technologies) and will be marked with a special label.
2.2. Personalization: this feed is not personalized and is displayed identically to all users.
2.3. User preference data used and its source:
data from the internal tracking system regarding page visit counts for specific Site pages over a given period.
2.4. Operational logic: the feed of 5 (five) publication links is generated automatically by ranking Site pages based on the number of users currently (or recently) on the page during a specific period preceding the feed generation.
3. Recommendations in the “Daily Habr” Feed
3.1. General description: a block containing links to 9 (nine) most interesting publications posted on the Site during the previous day.
3.2. Personalization: this feed is not personalized and is displayed identically to all users.
3.3. User preference data used and its source:
statistical data from the internal tracking system regarding user interaction with publication pages, such as visit counts, time spent on the page, scroll depth (percentage and amount), and others;
data from the internal tracking system on active user reactions to publications using tools available on the Site, such as number of comments, number of bookmarks, number of user votes (rating), and others.
3.4. Operational logic: the feed is generated automatically by ranking publications posted on the Site over the previous day using a mathematical algorithm that takes into account the data specified in paragraph 3.3 of these Rules, as well as the type of publication and the stream in which it is placed.
4. Recommendations in the “Daily best” Feed
4.1. General description: a block containing links to 20 (twenty) publications posted on the Site over the last 24 hours with the highest rating. A link to another publication may be added manually (without using recommendation technologies) and will be marked with a special label.
4.2. Personalization: this feed is not personalized and is displayed identically to all users.
4.3. User preference data used and its source:
data from the internal tracking system regarding the number of user votes for and against a publication (rating).
4.4. Operational logic: the feed is generated automatically by ranking publications posted on the Site over the last 24 hours based on the number of user votes for and against each publication (rating).
5.Recommendations in the “Similar” Feed
5.1. General description: a block containing links to 5 (five) publications similar to the publication the user is currently viewing.
5.2. Personalization: this feed is personalized and takes into account data about the user viewing a specific publication.
5.3. User preference data used and its source:
data from the internal tracking system regarding the user's viewing of a particular publication.
5.4. Operational logic: the feed is generated automatically by matching the publication being viewed by the user with other publications posted on the Site. The selection uses a machine‑learning algorithm that compares the vector of the current publication and searches for similar publications through their vectors. The publications identified in this manner are then ranked by rating and displayed in the feed.
6. Advertising Banners
6.1. Advertising banners may be displayed on the Site using the Adfox platform, which employs its own recommendation technologies. The rules for their use are available at: https://yandex.ru/legal/recommendations/ru/#index__adfox.