Functional Cluster-Based Ad Targeting for Mobile Apps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing software applications often provide multiple functions, but users typically utilize only a subset of these, leading to inefficiencies in advertising and search result relevance, as current systems fail to effectively target specific user interests based on application states.
Innovation Solution
A deep-linking system that includes an analytics engine for clustering software application states based on usage data and an advertising engine that generates sponsored links by identifying clusters and determining advertisement scores, ensuring that ads are targeted to users based on their likely interests and usage patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If software applications provide multiple functions to users, then the application offers greater versatility and utility, but users typically utilize only a subset of these functions, leading to inefficiencies in advertising and search result relevance
Solution Approach 1:
The patent segments application states into functional clusters based on usage patterns. Instead of treating all application functions as a single undifferentiated unit, the system divides them into distinct clusters (e.g., navigation cluster, content consumption cluster, communication cluster) that reflect actual user behavior patterns. This segmentation enables targeted advertising and search results specific to each cluster, resolving the contradiction between providing versatile functions and maintaining advertising relevance.
Solution Approach 2:
The system changes the parameter of state representation from generic application states to clustered functional states. By transforming the way application states are categorized and represented, the system can better capture user interests even when users only engage with a subset of functions. This parameter change enables more precise matching between user behavior and advertising content.
2Reliability
If the system targets advertising to specific user interests based on application states, then advertising effectiveness improves, but the complexity of identifying and processing application state clusters increases
Solution Approach 1:
The patent creates a universal clustering framework that can be applied across multiple applications and devices. The functional cluster identification system uses standardized criteria and data structures that work consistently regardless of the specific application or device type. This universality reduces the complexity burden by providing a repeatable, scalable approach rather than requiring custom solutions for each application.
Solution Approach 2:
The system enables applications and devices to automatically identify and report their own functional clusters based on usage data, without requiring manual configuration or complex external analysis. The clustering algorithm processes usage patterns autonomously to generate relevant clusters, reducing the operational complexity of implementing targeted advertising.
3Measurement precision
If the system processes and analyzes usage data to identify functional clusters, then user-specific advertising relevance improves, but the computational resources and processing time required increase
Solution Approach 1:
The system implements partial processing by focusing computational resources on identifying the most salient functional clusters rather than analyzing every possible application state in detail. The clustering algorithm prioritizes processing usage data that provides the most information about user interests, using less computational power for states with lower informational value. This selective processing maintains measurement precision while reducing energy consumption.
4Measurement precision
If the system generates sponsored links based on functional clusters, then search result relevance improves, but the time required to process advertisement requests and generate scores increases
Solution Approach 1:
The system performs preliminary clustering of application states in advance, creating a ready-to-use framework of functional clusters before advertisement requests arrive. This pre-processing establishes the structural foundation for rapid advertisement scoring and matching. When advertisement requests are received, the system only needs to evaluate which pre-defined clusters are relevant, significantly reducing processing time while maintaining high relevance through the pre-established cluster structure.
Data Source
AI summary
A deep linking system includes an advertising engine. The advertising engine includes one or more processors configured to receive an advertisement request and identify one or more advertisement records based on the advertisement request and triggering data included in the one or more advertisement records. The one or more processors are further configured, for each advertisement record, to identify one or more cluster records based on the underlying sponsored state of the identified advertisement record and determine an advertisement score for the identified advertisement record based on the identified cluster records. The one or more processors select one or more of the identified advertisement records based on the advertisement scores, generate advertisement objects based on the selected advertisement records, and transmit the advertisement objects to the remote device.