Helper Behavior–Driven Supplemental Content Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing advertising methods rely on past user behavior and fail to leverage new knowledge, resulting in irrelevant ads that do not account for the user's current needs or preferences.
Innovation Solution
A system that tracks a helper's interactions during a conversation with a user, identifies relevant products or services, and uses the helper's knowledge to provide supplemental content that aligns with the user's interests, leveraging natural language processing, artificial intelligence, and machine learning algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If search engine recommendation approaches are used to select ads based on user's past browsing history, then ads can be targeted to user interests, but the approach only looks at past history and does not leverage new knowledge from helper interactions
Solution Approach 1:
The system performs preliminary tracking of the helper's interactions and behavior during the conversation before ad selection occurs. This allows the system to capture new knowledge about user needs that emerges during the helper-assisted search, rather than relying solely on historical data. The helper's subsequent actions (clicks, selections, modifications) are recorded and analyzed to inform ad recommendations.
Solution Approach 2:
The system implements a feedback loop where the helper's interactions with search results are continuously monitored and fed back into the ad selection process. The helper's behavior (what they click, what they ignore, what they modify in the search query) provides real-time feedback about user needs, which is then used to dynamically adjust and refine ad recommendations during the session.
2Productivity
If ads are selected based on ad words corresponding to products including competitors' products, then ads can be displayed when user searches for matching words, but the ads are not specifically based on user's current needs and only assume needs have not changed
Solution Approach 1:
The ad selection system transitions from a static, history-based approach to a dynamic, real-time approach. Ads are no longer selected solely based on predetermined ad words and historical data, but are dynamically adjusted based on the helper's current interactions and the evolving context of the user's needs during the conversation. The system adapts ad recommendations as new information becomes available.
Solution Approach 2:
The system changes the parameters used for ad selection from fixed historical metrics to dynamic real-time indicators. Instead of relying on static browsing history and ad word matching, the system incorporates changing parameters such as the helper's current selections, modifications to search queries, and real-time interaction patterns, allowing ad relevance to evolve with user needs.
3Measurement precision
If methods ask the user what is best for them to determine the best product, then user preferences can be directly obtained, but this interferes with user experience and users may not know what is out in the marketplace
Solution Approach 1:
The system introduces the helper as an intermediary between the user and the ad selection process. Instead of directly querying the user about their preferences, the system observes and analyzes the helper's interactions, selections, and modifications during the search process. The helper's expert knowledge and behavior serve as a mediator to infer user needs without disrupting the user experience.
Solution Approach 2:
The system enables self-service by automatically inferring user preferences through passive observation of helper behavior rather than requiring active user input. The analysis engine autonomously processes the helper's interactions and generates ad recommendations based on inferred user needs, eliminating the need for users to manually express their preferences or interrupt their workflow.
Data Source
AI summary
Systems and methods for selecting supplemental content based on a conversation between a user and a helper are disclosed. The methods analyze a conversation, or other interactions, to determine the context and determine whether content discussed is associated with a product or service in the marketplace and of interest to the user. A factor-based determination is made whether to track helper's behavior and interactions after the conversation. When the helper is tracked, based on their tracing preferences, the helper's biometrics, such as gaze and/or heartbeat, are used to determine their interest in a product or service consumed, such as via media consumption, virtual reality experiences, or other online and offline interactions. Supplemental content related to helper's interested products are filtered based on user preferences and transmitted to the user. Knowledge trees associated with the supplemental content that include the basis for recommendation, source of knowledge, and other knowledge related details are provided for user access.