Digital Asset Inventory Precomputation for Low-Latency Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for constructing digital asset inventories for organizations are slow and inefficient, particularly for large organizations, due to the high latency of recursive query processes and the difficulty in incorporating additional data sources, which hinders real-time risk assessment and security performance evaluations.
Innovation Solution
A method and system that models digital asset relationships as a graph structure, enabling low-latency retrieval by preprocessing and indexing digital assets, using bipartite graphs and graph algorithms to filter and curate data, and incorporating blocklists and human curation to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a recursive query process is used to build digital asset inventory, then the inventory can be constructed from known assets, but the process becomes too slow for large organizations and real-time requests
Solution Approach 1:
The patent precomputes and stores the complete digital asset inventory in a database before queries are executed. By performing the asset discovery and relationship mapping in advance (preliminary action), the system eliminates the need for slow recursive queries at request time, enabling real-time retrieval while maintaining comprehensive inventory accuracy.
Solution Approach 2:
The patent creates a static copy of the complete digital asset inventory and stores it in a database. Instead of performing recursive queries on live data sources each time, the system uses a pre-generated copy of the inventory data, which dramatically reduces query latency while preserving the completeness of asset information.
2Quantity of substance
If additional data sources are incorporated into the inventory building process, then the comprehensiveness of the inventory improves, but the time to discover assets across all methods increases
Solution Approach 1:
The patent incorporates multiple data sources (DNS records, SSL certificates, WHOIS data, etc.) during the precomputation phase. By performing all data source queries and integrations in advance, the system achieves comprehensive asset coverage from multiple sources without incurring the time penalty during actual inventory requests.
Solution Approach 2:
The patent merges data from multiple sources (DNS, SSL, WHOIS, passive DNS, etc.) into a unified precomputed inventory structure. By combining all data source results during the precomputation phase and storing them in an integrated format, the system achieves comprehensive coverage while enabling fast retrieval through a single database query operation.
3Ease of manufacture
If a recursive query approach is used, then the inventory can be built systematically, but it becomes difficult to include additional data sources without further lengthening the time
Solution Approach 1:
The patent performs all systematic inventory construction activities in advance, including integrating multiple data sources. The precomputation phase systematically processes all available data sources and stores the results in a structured database, making the system both easy to construct and highly adaptable to additional sources without affecting query performance.
Solution Approach 2:
The patent creates a universal database schema that can accommodate multiple types of data sources (DNS, SSL, WHOIS, passive DNS, etc.). The precomputed inventory structure is designed to be multi-functional, allowing systematic organization of diverse data while enabling fast retrieval regardless of which data sources were used during precomputation.
Data Source
AI summary
Methods and supporting systems for discovering and attributing computational assets to respective entities includes receiving datasets comprising source record data for numerous of entities and including attributes such as point of contact data elements and normalizing the data elements to maintain a consistent format across the source record data. A filter is applied against the normalized source record data to identify data that are either (i) redundant or (ii) refer to a bulk control entity. A further curation step can be applied against the data as the data is stored as a bipartite graph, wherein a first set of nodes of the bipartite graph represents source records and a second set of nodes represents attributes of the source records.