Geo-temporal Indexing via Temporal Buckets for Low-Latency Query Processing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current geospatial indexing techniques are inefficient for processing geo-temporal data due to the added dimensionality of time, leading to high latency in searches and complications in indexing data with varied temporal granularities.

Innovation Solution

A geo-temporal indexing structure is generated and managed in computer memory, incorporating both geospatial and temporal components, using techniques like Z-order curves and Hilbert curves, with user-specified time granularities to create temporal index buckets that facilitate efficient querying.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional geospatial indexing techniques are used on geo-temporal data, then geographical locality is preserved in the indexed data, but processing efficiency deteriorates due to the added dimensionality of time

Engineering Contradiction:
Improvegeographical locality preservationVSAvoidquery processing efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent segments the geo-temporal index into multiple temporal buckets, each representing a specific time range. This segmentation allows the system to preserve geographical locality within each bucket while enabling efficient time-based filtering at the bucket level, thus resolving the contradiction between maintaining spatial relationships and achieving efficient query processing on multi-dimensional geo-temporal data

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a temporal dimension to the traditional geospatial index by organizing data into temporal buckets. This dimensional extension allows the index to handle geo-temporal queries efficiently by first filtering on the temporal dimension (bucket level) and then applying geospatial operations only to relevant subsets, thereby maintaining geographical locality while improving productivity

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Reliability

If geospatial indexing mechanisms are applied to geo-temporal data, then data can be organized by geographical location, but search latency increases due to inefficient handling of temporal variations

Engineering Contradiction:
Improvegeospatial data organizationVSAvoidsearch latency
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent performs preliminary organization of geo-temporal data into temporal buckets during the indexing phase. By pre-grouping data according to time ranges and maintaining geospatial structure within each bucket, the system eliminates the need for complex temporal filtering during query execution, thus preserving geospatial organization while significantly reducing search latency

Inventive Principle:
Principle #10Preliminary action

3Adaptability or versatility

If data is indexed with varied temporal granularities, then user queries with different time resolutions can be supported, but indexing complexity increases

Engineering Contradiction:
Improvetemporal granularity supportVSAvoidindexing structure complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent implements segmentation by creating multiple temporal buckets with different granularity levels (e.g., hourly, daily, monthly buckets). Each bucket maintains a simplified geospatial structure, avoiding the need for a single complex indexing structure to handle all temporal granularities. This segmented approach supports versatile temporal queries while keeping individual bucket structures relatively simple

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The temporal bucket structure serves multiple functions: it organizes data by time ranges, enables efficient temporal filtering, maintains geospatial locality, and supports various temporal granularities. This multi-functional design achieves adaptability without proportionally increasing complexity, as the same basic bucket structure handles diverse temporal query requirements

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS10691662B1Geo-temporal indexing and searching
Publication Date: 2020.06.23 PALANTIR TECHNOLOGIES INC
  • US10691662B1 patent drawing
  • US10691662B1 patent drawing
  • US10691662B1 patent drawing

AI summary

A method and apparatus for a data analysis system for analyzing data object collections that include geo-temporal data is provided. One or more temporal granularities are specified for the purpose of generating a geo-temporal data index. The time granularities correspond to temporal ranges expected to correspond to temporal ranges specified in user queries against the data. One or more temporal index bucket groups are generated based on to the specified time granularities. Geo-temporal input data is indexed based on the generated temporal index bucket groups. The system allows a data analyst to specify geo-temporal queries that include both geospatial component and a temporal component. The system transforms geo-temporal queries into one or more second queries that retrieve data items based on the temporal index bucket groups.