Affinity Engine®

Affinity Engine® – Pat. Pending 

The CliqRex Inc., Affinity Engine® is powered by user ratings compared to other users’ ratings. Affinity is scored from 0-100 value. 0 being no affinity and 100 as perfect affinity.  

The following calculations are used to determine the affinity rating:

Closeness of rating value. This input signal is based on the difference between items users have both rated. Closer values = higher affinity

Love and Hate 
Our 11-point likert scale is far more powerful than thumbs up and thumbs down or five-star system as it gives the user the choice of 11 data points (0-10) vs. 2 (up or down) or 5 (1-5 stars). This creates far more variance in the data which results in more accurate and robust measures of similarity and difference.

AI – Sentiment Analysis 

This input is based on users’ comments and similar feelings on the rated item. Emoji analysis as well. Uses Google’s BERT A.I. language model.

Consumption similarity. The common items that you share in rating vs. the overall items you consume (took the time to rate) 

Accuracy or RexFactor

The percentage shows how alike your taste is, based on items you have both rated. The percentage is visualized by how complete the ring is around the profile pic. 

You could have a high-affinity percentage, but the ring is very faint or not visible at all. Transparent rings need more information to be considered accurate. If you both rate more of the same items, it will brighten the ring, creating a more accurate reflection of your affinity. 

Weaker Affinity Calculation – the ring is more transparent

Quickly see how much you have in common with another user by glancing at their Affinity RingTM

The colorful ring around other users is the Affinity Ring. The more the ring is complete, the more your tastes and ratings are similar. Transparent rings need more information to be considered accurate. Rate more items to help calculate your affinity score. 

The affinity ring allows Users to quickly visualize their affinity match with another user based on the completeness of the affinity ring around a user’s profile image. The more the ring is complete, the more the user’s tastes are aligned. Higher affinity = greater similarity in taste. The ring’s transparency indicates how many data points were available to be calculated. The more the Users interact with the CliqRex Application the greater accuracy of the Affinity Algorithm.