Trend Leader Detection in Product Recommendation Systems

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional information processing systems fail to effectively recommend products based on purchasing behavior that creates trends, leading to insufficient promotion of sales and inability to acquire information on leading degrees of users or trend shifts using behavior history.

Innovation Solution

An information processing system with a server apparatus and output apparatus that detects trend leaders by analyzing purchase history, recommending products purchased by these leaders, and promoting sales by identifying and recommending trendy products and services.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional information processing systems are used for content recommendation, then content recommendation based on user interest and sequentiality can be achieved, but the system cannot identify trend leaders or capture trend shifts in purchasing behavior

Engineering Contradiction:
Improverecommendation accuracyVSAvoidtrend information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent segments users into trend leaders and trend followers based on their purchasing behavior patterns. By dividing the user base into these distinct groups, the system can analyze and recommend products differently for each segment, capturing trend information that was previously lost in conventional unified recommendation approaches.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs preliminary analysis of purchasing behavior to identify trend leaders before making recommendations. By pre-identifying users who exhibit trend-leading behavior (purchasing items before others), the system can proactively capture trend shifts and use this information to improve recommendation accuracy for both trend leaders and followers.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If purchase history information is analyzed to identify trend leaders, then product recommendation effectiveness improves, but system complexity increases

Engineering Contradiction:
Improvesales promotion effectivenessVSAvoidsystem structure
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The system uses purchase history information that is already being collected and stored by the e-commerce platform. Instead of requiring additional external data sources or complex external analysis systems, the platform leverages its existing data infrastructure to identify trend leaders and generate recommendations, thereby improving sales promotion effectiveness without proportionally increasing system complexity.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent merges the trend leader identification function with the existing content recommendation system. By combining the analysis of purchase history patterns with the recommendation generation process, the system achieves improved productivity in sales promotion while avoiding the need for separate, complex systems. The trend leader identification is integrated into the existing architectural framework.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentUS8700489B2Apparatuses and method for recommending items based on determined trend leaders and trend followers
Publication Date: 2014.04.15 SCIGINEER
  • US8700489B2 patent drawing
  • US8700489B2 patent drawing
  • US8700489B2 patent drawing

AI summary

A server apparatus includes: a trend leader detecting portion that acquires purchaser identifying information for identifying a purchaser who purchased the product at an early stage satisfying a predetermined time condition, from at least two pieces of purchase history information; a recommended product acquiring portion that acquires at least one piece of product identifying information for identifying a product purchased by at least one purchaser identified with at least one piece of purchaser identifying information that has been acquired by the trend leader detecting portion; and a recommended product transmitting portion that transmits the at least one piece of product identifying information that has been acquired by the recommended product acquiring portion, to an information output apparatus. With this server apparatus, recommendation of a product can be provided based on the behavior in which a trend leader purchases a product.