N-gram Trend Detection for Online Marketplace Fraud
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems fail to accurately identify trending items in online marketplaces that are likely to be involved in fraudulent transactions, leading to increased risk and financial losses.
Innovation Solution
A computer program decomposes item descriptions into N-grams, monitors their frequency over time, and applies thresholds to detect trending items, which are then subjected to heightened scrutiny to prevent potential fraud.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional systems monitor all transactions equally, then transaction security is maintained, but fraud detection accuracy deteriorates because trending items cannot be identified
Solution Approach 1:
The patent segments transactions into different risk categories based on item trend analysis. By decomposing the monitoring system into segments that specifically target trending items versus normal items, the system can apply differentiated scrutiny levels. This segmentation allows enhanced fraud detection for high-risk trending transactions while maintaining efficient processing for routine transactions, thereby improving overall detection accuracy without compromising security.
Solution Approach 2:
The patent implements preliminary action by proactively identifying trending items before fraudulent transactions occur. The system continuously monitors item popularity and trend patterns, flagging items that are gaining traction. This preliminary identification allows the system to pre-establish heightened monitoring protocols for these items, enabling faster and more accurate fraud detection when suspicious transactions involve trending items.
2Reliability
If heightened scrutiny is applied to all transactions, then fraud prevention improves, but system efficiency deteriorates due to increased processing overhead
Solution Approach 1:
The patent applies local quality by implementing differentiated monitoring intensity based on item characteristics. Transactions involving trending items receive heightened scrutiny with more rigorous validation checks, while transactions involving non-trending items undergo standard processing. This localized quality adjustment ensures that fraud prevention resources are concentrated where they are most needed, maintaining high reliability for risky transactions while preserving system efficiency for routine operations.
Solution Approach 2:
The patent employs partial action by applying heightened scrutiny only to a subset of transactions involving trending items rather than all transactions. The system identifies specific high-risk categories and applies enhanced monitoring protocols selectively to these areas. This partial application of excessive scrutiny maintains strong fraud prevention capabilities for vulnerable transactions while avoiding the efficiency penalties of universal intensive monitoring.
3Measurement precision
If automated trend detection is implemented, then fraud identification capability improves, but system complexity increases due to N-gram processing requirements
Solution Approach 1:
The patent implements self-service by enabling the system to automatically perform trend detection and item identification without requiring external manual analysis. The N-gram processing pipeline autonomously scans item descriptions, extracts meaningful phrases, tracks their frequency over time, and identifies trending items. This self-service capability maintains high trend detection accuracy while managing complexity through automation, reducing the need for manual intervention in the monitoring process.
Solution Approach 2:
The patent replaces manual fraud analysis mechanics with automated computational systems. Instead of relying on human analysts to identify trending items through manual review, the system uses algorithmic N-gram processing and statistical analysis to automatically detect trends. This substitution of mechanical human analysis with automated computational mechanisms maintains high detection precision while managing system complexity through efficient algorithmic processing.
Data Source
AI summary
Descriptions of items are offered for sale by one or more merchants in an online marketplace. The online marketplace comprises a website hosted by a server. The descriptions are electronically scanned or otherwise accessed. The scanned descriptions are deconstructed into a plurality of N-grams. Each N-gram includes a combination of words appearing in the descriptions of items. The electronic scan and deconstruction are repeated over a plurality of predefined time periods. For each N-gram, a frequency of occurrence is monitored in each of the predefined time periods. Based on the monitoring, a determination is made that a first N-gram of the plurality of N-grams whose frequency of occurrence has exceeded a predefined threshold in one of the predefined time periods. Risks of transactions involving one or more items whose descriptions contain the first N-gram are evaluated.