Intent Chain Prediction for Online Ad Relevance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current advertisement methods primarily rely on direct keyword matches between online content and ads, failing to anticipate user intentions beyond their immediate actions, thus missing opportunities for relevant but non-consciously intended content to engage users.
Innovation Solution
A system that determines user intentions through analyzing sequences of online actions, creating 'chains of intent' to predict related future intentions, allowing for the display of online objects like ads that align with these anticipated interests, even if not directly related to the user's current action.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If advertisements are presented based on direct keyword matches between online content and ads, then ad relevance to current content is improved, but user engagement and anticipation of future intentions are lost
Solution Approach 1:
The system performs preliminary analysis of user behavior patterns and intent chains before the user actually needs the advertisement. By predicting future intentions based on sequences of past actions (e.g., researching cars → buying car insurance), the system prepares and presents ads proactively rather than reactively, resolving the contradiction between current relevance and future engagement
Solution Approach 2:
The advertisement system transitions from static keyword matching to dynamic intent prediction. The system continuously updates user intent models based on evolving behavior patterns, allowing ads to adapt to changing user needs in real-time. This dynamic approach enables the system to maintain relevance while anticipating future intentions, thereby improving both ad relevance and user engagement
2Ease of operation
If advertisements are presented based on explicit user intentions, then user control is improved, but opportunities for serendipitous engagement are lost
Solution Approach 1:
The system introduces an intermediary layer of intent prediction between user actions and advertisement presentation. Instead of directly responding to explicit intentions or showing all possible ads, the intermediary intent chain analysis filters and selects ads that are likely to resonate based on inferred patterns. This mediator enables serendipitous engagement by suggesting relevant content the user didn't explicitly seek but would likely appreciate
Data Source
AI summary
Methods, systems, and computer-readable media for identifying intentions related to a user's present intention are provided. The user's present intention may be determined by mapping the user's current actions to a present intent. A related intent may be determined using an intent chain. The intent chain may be generated by analyzing patterns of user activities and mapping the patterns of activities to the underlying intents.