Mini project "Vizitator" - user preferences supervisor
On the wave of positive feedback on the article about data collection through CSS, I really wanted not to postpone the idea to the far corner, but quickly implement the final product using it. As a result, a mini-service Vizitator was created, about which I will tell respected habrap users.
Vizitator is a javascript API (although so far it is said loudly), which allows to obtain data about the user's interests based on his preferences for site visits.
JS inserts link arrays on the page and analyzes their color, creating two different styles for normal and: visited links in advance. Since the browser applies colors when the link appears on the page - it is possible to immediately determine whether the user was at this address or not.
')
The current release is rewritten on bare js and minimized (jQuery is good, but alas, too much to drag along), and slightly reworked relative to the original scripts (links are inserted and analyzed one by one).
Online shopping and social networks - you can show new users the title (or any other) pages sharpened just for their interests.
Banner networks - you can make ads relevant to the interests of users (and not just the content of the pages, although I suspect that they already use similar technologies)
Any sites with recommendation services.
PR - service may allow to monitor visits to individual links on any sites.
Connect the js file to your website, either by downloading it to yourself or by placing a link to the original (this will allow you to automatically receive updates of the database). Define callback function
and within it respond to the data about preferences (the help has not yet been written in detail, so the example is the service page itself). Vizitator is activated by the page loading event and works almost instantly.
The keys for use will appear in the next version, as long as everything is completely open to everyone, but when the beta version is released, the current one will be removed.
Now the project base is about 160 links, which is quite a bit, it is planned to increase the base by 3-4 times, add new categories and make the analysis multi-level (that is, limit it to a test set of 100-200 positions, and load the rest via ajax). Perhaps jQuery plugin to further reduce the size for those who use jQuery. And also add. user data collection services.
In fact, we are good and fluffy (and not evil monsters, cookie hijackers), one of the goals of the service is to draw attention to the problem in order to protect users from less friendly people. On the other hand, the service only provides a functional base, and each integrator should decide for itself whether to issue any warnings to the user or immediately collect its data.
At the moment, the project is in the alpha stage, which means that we are deciding whether there will be any benefit from it, or is it a waste of time.
We are open to proposals for the integration of the project, if you are interested in the idea - write, invent and implement.
Vizitator is a javascript API (although so far it is said loudly), which allows to obtain data about the user's interests based on his preferences for site visits.
How it works (for those who are lazy to read the article)
JS inserts link arrays on the page and analyzes their color, creating two different styles for normal and: visited links in advance. Since the browser applies colors when the link appears on the page - it is possible to immediately determine whether the user was at this address or not.
')
The current release is rewritten on bare js and minimized (jQuery is good, but alas, too much to drag along), and slightly reworked relative to the original scripts (links are inserted and analyzed one by one).
Who is useful
Online shopping and social networks - you can show new users the title (or any other) pages sharpened just for their interests.
Banner networks - you can make ads relevant to the interests of users (and not just the content of the pages, although I suspect that they already use similar technologies)
Any sites with recommendation services.
PR - service may allow to monitor visits to individual links on any sites.
How to use
Connect the js file to your website, either by downloading it to yourself or by placing a link to the original (this will allow you to automatically receive updates of the database). Define callback function
VizitatorCallback(keys, values)
and within it respond to the data about preferences (the help has not yet been written in detail, so the example is the service page itself). Vizitator is activated by the page loading event and works almost instantly.
The keys for use will appear in the next version, as long as everything is completely open to everyone, but when the beta version is released, the current one will be removed.
Development plans
Now the project base is about 160 links, which is quite a bit, it is planned to increase the base by 3-4 times, add new categories and make the analysis multi-level (that is, limit it to a test set of 100-200 positions, and load the rest via ajax). Perhaps jQuery plugin to further reduce the size for those who use jQuery. And also add. user data collection services.
Privaci
In fact, we are good and fluffy (and not evil monsters, cookie hijackers), one of the goals of the service is to draw attention to the problem in order to protect users from less friendly people. On the other hand, the service only provides a functional base, and each integrator should decide for itself whether to issue any warnings to the user or immediately collect its data.
Request to cooperate
At the moment, the project is in the alpha stage, which means that we are deciding whether there will be any benefit from it, or is it a waste of time.
We are open to proposals for the integration of the project, if you are interested in the idea - write, invent and implement.
Source: https://habr.com/ru/post/46784/
All Articles