User Activity Feature Vector Mapping for Item Association

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional methods for associating user activity indications with items, such as banners or products, are cumbersome and inaccurate, requiring manual keyword selection and lacking automation, which hinders efficient data utilization and machine learning applications.

Innovation Solution

A computer-implemented method that generates activity feature vectors from user activity indications, combines them with text and image feature vectors to create a multidimensional space, reflecting associations between user activities and items, thereby automating the association process and improving accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual keyword selection is used to associate items with user activity indications, then the association process can be performed, but the process becomes cumbersome and inaccurate

Engineering Contradiction:

Improveaccuracy of item associationsVSAvoidoperational complexity

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The system automatically generates keywords and associations between items and user activity indications without requiring manual intervention. The computer-implemented method autonomously processes user activity data, extracts meaningful patterns, and creates associations, thereby eliminating the need for manual keyword selection and significantly improving both accuracy and operational efficiency

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces the manual mechanical process of keyword selection with an automated computational system. The system uses algorithms to process user activity indications, generate relevant keywords automatically, and establish item associations through data-driven approaches rather than human judgment, thereby improving precision while reducing operational burden

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If manual keyword tagging is performed for each item, then item associations can be created, but the process is not cost-efficient and results in inaccuracies

Engineering Contradiction:

Improveaccuracy of associationsVSAvoidcost-efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system performs self-service by automatically generating keywords and associations without human intervention. The computer-implemented method processes user activity indications, autonomously identifies relevant items, and creates accurate associations through algorithmic analysis, thereby improving reliability while eliminating the costs and time associated with manual tagging operations

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system creates automated copies of the association process by using algorithms to replicate the function of manual keyword selection. Instead of requiring human experts to manually tag each item, the system generates keyword associations automatically through computational methods, improving both accuracy and cost-efficiency by scaling the process without additional human resources

Inventive Principle:

Principle #26Copying

3Adaptability or versatility

If conventional methods are used to process user activity data, then basic associations can be made, but the system cannot capitalize on previous associations or improve through machine learning

Engineering Contradiction:

Improvelearning capabilityVSAvoidsystem complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system implements feedback mechanisms by analyzing user interactions with previously associated items and using this information to improve future associations. The computer-implemented method continuously processes user activity indications, learns from patterns in the data, and refines keyword generation and item association algorithms, thereby enabling the system to adapt and improve over time while managing complexity through systematic data-driven approaches

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system performs preliminary actions by pre-processing user activity data and pre-generating keyword associations before they are needed for specific matching tasks. The computer-implemented method anticipates association needs by continuously analyzing user behavior patterns and preparing relevant item associations in advance, thereby enabling faster and more accurate real-time recommendations without overwhelming system complexity during critical matching operations

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11157522B2Method of and system for processing activity indications associated with a user

Publication Date: 2021.10.26 Y E HUB ARMENIA LLC

AI summary

A method (1400) of and a system (222) for associating past activity indications (602) associated with past activities of a user (170) with items. The method comprises accessing (1402) the past activity indications (602); accessing (1404) item indications; determining (1406) a past activity feature vector (606); determining (1408) a text feature vector (706) corresponding to the text features; mapping (1410) the past activity feature vector (606) and the text feature vector (706) to generate a text feature space (904); determining (1412) an image feature vector (806); mapping (1414) the past activity feature vector (606) and the image feature vector (806) to generate an image feature space (1004); generating a user item space (1104); and storing (1418) the user item space (1104). A method (1500) of and a system (222) for associating a first item and a second item are also disclosed.