Digital Item Recommendation via Desirability Scoring and Ranking

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional digital content distribution systems fail to recommend the most desirable items to users as they do not consider the desirability of items, leading to suboptimal user experience.

Innovation Solution

A method that computes a desirability score for each digital item based on scoring metrics, ranks them relative to others, and evaluates them to determine group definitions, selecting the most desirable items and their group definitions for recommendation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional recommendation systems select items based on group definitions from user's previously selected items, then the system can provide personalized recommendations, but the system fails to consider the desirability of items leading to suboptimal recommendations

Engineering Contradiction:

Improveitem desirability assessmentVSAvoidrecommendation quality

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent introduces a desirability score parameter that quantifies item attractiveness based on user interaction metrics (plays, skips, shares). This parameter transformation enables the system to objectively assess and prioritize items, resolving the contradiction between personalized grouping and recommendation quality by adding a measurable dimension to item evaluation.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces the conventional mechanical approach of simple frequency counting with a scoring mechanism that incorporates multiple interaction metrics. This substitution transforms the recommendation system from a basic categorization tool to an intelligent evaluation system that can distinguish between desirable and less desirable items within the same group.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If the system evaluates all digital items to determine group definitions, then comprehensive recommendations can be provided, but the computational complexity and time required increase significantly

Engineering Contradiction:

Improverecommendation accuracyVSAvoidevaluation time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent performs preliminary ranking of items based on desirability scores before evaluating them for group definition. This preliminary action filters and prioritizes items, allowing the system to focus evaluation resources on the most promising candidates rather than processing all items equally, thus maintaining reliability while reducing time loss.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent segments the evaluation process into distinct stages: first computing desirability scores for all items, then ranking them, and finally evaluating items in order of rank to determine group definitions. This segmentation allows the system to process items efficiently in prioritized batches, balancing comprehensive evaluation with computational efficiency.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS8903834B2Recommending groups of items based on item ranks

Publication Date: 2014.12.02 NETFLIX INC

AI summary

One embodiment of the invention sets forth a mechanism for recommending digital items to a user. Each digital item in a set of digital items is scored based on user preferences and other metrics. The digital items are ordered based on scores. The digital items are then evaluated in order of respective rank to determine a subset of digital items that re recommended to the user. The evaluation process is based on different evaluation criteria as well as the presentation style of the recommended digital items.