Distributed Indexing for Social Media Query Processing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Social media networks face significant computational challenges in processing and searching large and dynamic datasets due to the prohibitive processing requirements, which limits real-time understanding and processing of vast amounts of data.

Innovation Solution

A distributed processing cluster utilizing a plurality of nodes, including a query proxy, graphics processing units, and various indexes (bitmap, free text search, interval, and log structured merge tree indexes), which executes queries and aggregations efficiently by using fixed width unique identifiers and stored fields to facilitate real-time querying and aggregation of social media datasets.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If a distributed processing cluster is used to process social media datasets, then processing efficiency and scalability are improved, but system complexity increases

Engineering Contradiction:
Improveprocessing efficiencyVSAvoidsystem complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent divides the social media dataset into multiple partitions distributed across different nodes in the cluster. Each node processes a specific partition independently, enabling parallel processing and improving overall efficiency while keeping individual node complexity manageable.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The query proxy serves as an intermediary between the user and the distributed processing cluster. It receives queries, determines execution plans, and coordinates the processing across multiple nodes, thereby managing system complexity while maintaining high processing efficiency.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Speed

If data is distributed across multiple nodes, then query execution speed is improved, but data management complexity increases

Engineering Contradiction:
Improvequery execution speedVSAvoiddata management complexity
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The dataset is segmented into partitions that are distributed across multiple nodes. This segmentation enables parallel query execution on different partitions simultaneously, improving query speed while maintaining manageable data management through structured partitioning.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The data source structure is designed to be universal and consistent across all nodes, with each node handling multiple operations (storage, processing, indexing). This multi-functionality simplifies data management by providing a uniform interface despite the distributed nature of the system.

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

3Speed

If fixed width unique identifiers are used, then processing speed is improved, but data storage flexibility decreases

Engineering Contradiction:
Improveprocessing speedVSAvoiddata storage flexibility
Core Design Contradiction:
SpeedVSAdaptability or versatility

Solution Approach 1:

The patent changes the parameter of identifier format to fixed width, which enables more efficient processing and comparison operations. This parameter change improves processing speed while the system compensates for reduced flexibility through other mechanisms like structured data sources and indexing.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS10902069B2Distributed indexing and aggregation
Publication Date: 2021.01.26 RUNTIME COLLECTIVE LTD
  • US10902069B2 patent drawing
  • US10902069B2 patent drawing
  • US10902069B2 patent drawing

AI summary

Data characterizing a query of a social media dataset can be received. The query can be executed utilizing a distributed processing cluster. The distributed processing cluster can include a plurality of nodes. At least one node can execute a first query on a partition of a tablespace storing a portion of the social media dataset. The partition can include a data source that can include a fixed width unique identifier and a stored field. The fixed width unique identifier can be associated with a respective record of the social media dataset and the stored field can include a portion of the respective record. A result of the query can be provided. Related apparatus, systems, techniques and articles are also described.