Adjusting Messaging Link Presentation Value Using Negative Experience Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Online systems face challenges in accurately predicting user affinities for content that includes links launching messaging applications, as they cannot determine whether negative experiences occurred during conversations initiated by these links, affecting the accuracy of content presentation values.
Innovation Solution
An online system adjusts the value associated with presenting a link that initiates a conversation by training a machine-learning model using information on negative experiences and entity attributes, predicting the likelihood of a negative experience, and adjusting the presentation value based on this prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the online system presents content items including links that launch messaging applications to users, then user engagement is maximized, but the accuracy of predicting user affinities for such content deteriorates because the system cannot determine whether negative experiences occurred during conversations
Solution Approach 1:
The system performs preliminary actions by collecting feedback information from users after messaging conversations occur. This feedback is then used to train a machine learning model that can predict the likelihood of negative experiences, enabling the system to make informed decisions about content presentation before actual user interactions occur.
Solution Approach 2:
The system implements a feedback mechanism where user experiences during messaging conversations are collected and fed back into the system. This feedback loop allows the machine learning model to be continuously trained and improved, enabling more accurate predictions of user affinities and negative experience likelihoods, which in turn improves content selection accuracy.
2Measurement precision
If the online system collects and processes information about negative experiences to improve prediction accuracy, then the accuracy of content selection is improved, but the system complexity increases due to implementing machine learning models and feedback collection mechanisms
Solution Approach 1:
The system introduces a machine learning model as an intermediary component that bridges the gap between raw feedback data and content selection decisions. This intermediary processes the complex feedback information and transforms it into actionable predictions about negative experience likelihood, enabling accurate content selection without requiring the entire system to become exponentially more complex.
Solution Approach 2:
The system implements self-service capabilities by automatically collecting feedback information, training the machine learning model, and applying predictions to content selection without requiring manual intervention. This automation reduces operational complexity while maintaining high prediction accuracy.
Data Source
AI summary
An online system presents content to its users, in which the content includes links that launch a messaging application and initiate conversations via the application. The system receives information indicating that negative experiences occurred during the conversations and may use this information and attributes of entities participating in the conversations to train a model to predict a likelihood of an occurrence of a negative experience. Upon determining an opportunity to present a user with a link that launches the application and initiates a potential conversation with an entity via the application, the system applies the model to predict a likelihood of an occurrence of the negative experience by the user during the potential conversation based on the entity's attributes. Based on the predicted likelihood, the system adjusts a value associated with presenting the link and passes the adjusted value to a process that selects content for presentation to the user.