DSP Bid Win Tracking via Time and Cache Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the context of real-time bidding (RTB) ad auctions, there is a challenge in accurately managing bid win data due to issues with duplicate or fraudulent notifications, which can lead to data integrity concerns and inefficiencies in tracking bid wins.
Innovation Solution
A computer-implemented method and device that processes advertisement bid win notifications by checking if the time information satisfies a predetermined criteria, and if so, verifies the identity information's presence in a cache and database, recording the notification if it's not already present, while discarding notifications that fail these checks to prevent duplicates and fraud.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If all bid win notifications are recorded without verification, then the quantity of bid win data is maximized, but data reliability deteriorates due to duplicate or fraudulent notifications
Solution Approach 1:
The system performs preliminary verification actions before recording bid win notifications. It checks time information against a predetermined criteria, verifies identity information against cache and database, and validates data integrity information before accepting the notification. This preliminary action prevents duplicate and fraudulent data from being recorded, thus maintaining data reliability while still capturing valid bid wins.
Solution Approach 2:
The system uses feedback mechanisms by checking identity information against previously stored data in cache and database. The verification process compares incoming notifications with historical records, and the results of these checks feed back into the decision to accept or reject the notification. This feedback loop ensures that only unique, valid bid wins are recorded, preventing duplicates and fraud.
2Reliability
If strict verification criteria are applied to all notifications, then data reliability is improved by filtering duplicates and fraud, but processing time increases due to multiple checks
Solution Approach 1:
The verification process is segmented into distinct stages: first checking time information against predetermined criteria, then verifying identity information against cache, and finally checking against database if needed. This segmentation allows the system to perform quick filters first (time check, cache check) before proceeding to more time-consuming operations (database queries), thus reducing overall processing time while maintaining reliability.
Solution Approach 2:
The system applies different verification intensities to different aspects of the notification based on local quality principles. Time information receives a predetermined criteria check, identity information is verified against cache and database, and data integrity information is validated. Each aspect receives appropriate verification depth, avoiding unnecessary processing while ensuring critical reliability checks are performed.
3Reliability
If time information criteria are applied to filter old notifications, then data reliability is improved by excluding stale data, but the quantity of accepted notifications decreases
Solution Approach 1:
The system changes the parameter of time information by applying a predetermined criteria that evaluates the recency or validity of the timestamp. Notifications with time information that does not satisfy the criteria (e.g., too old, inconsistent timing patterns) are filtered out. This parameter-based filtering improves data reliability by excluding stale or suspicious notifications while maintaining a substantial quantity of valid, timely bid win data.
Data Source
AI summary
A demand side platform DSP is provided which will receive an advertisement bid win notification comprising identity information and time information, The DSP will check the time information to make sure the notification is not too old. If not the DSP will check if the identity information is in a cache. If not, a check is made against a database. If not, the bid win is recorded in the cache and the database.