Cases Of Use
Item recommendations based on direct measurements
This kind of interest measurement on the products is one of the most commons on some streaming applications, online stores, etc.
For instance, you could add a rating system
based on a number of stars, like a movies classification based on a 0 to 5 stars system. Based on the previous user classifications for the movies we will be able to predict what are the best movies for each user.
This is an easy and direct way, but sometimes the users lie about its preferences or prefer not to classify some controverted items. In order to improve the recommendations for this cases you can use indirect measurements.
Item recommendations based on indirect measurements
If you don't have any rating system implemented in your site, don't worry you can use any kind of indirect measurement of the user interest for the items, usually this kind of measurement is more precise than a direct measurement, behavior never lies
This kind of actions can determine the interest of a user in the product:
Time spent on the products Send to us the time spent by the users watching each of the products during the last week, and we will be able to predict what other products can be of interest for this user
Number of visits to the products Another sign of interest for an item could be the number of times that a user has visited it, so we can predict the interest of the user for items based on this information
Actions on a product If the user opens the attached images of the product, access to the full description of an article and so on, that could be an indicative of the grade of interest on the product, you can assign a score for each possible action, and send to us the sum of this scores by item
... Any other measurable indicator of interest can be analysed by Pitia in order to provide a great user experience to your customers!
Take care of the bots, a bot like Google, Bing, etc can have an arbitrary behaviour that doesn't give any value to the collected data, try to avoid sending this information to Pitia.