Parent Document Indexing for Unified Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-relational databases like Elasticsearch do not support joins between different document types, limiting the ability to query related documents such as tickets and their comments in a single search operation.
Innovation Solution
The solution involves linking parent and child documents within a parent document index by using identifiers, where the parent document is modified to include data from related child documents during indexing, allowing for robust search functionality across both parent and child data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If documents of different types are stored in separate indexes, then data organization is simplified and indexing is easier, but the ability to query related documents together is lost
Solution Approach 1:
The patent embeds child document data within parent document fields during indexing. The parent document index contains nested structures that hold child document information, allowing the system to maintain separate physical indexes while enabling unified queries through the nested data organization
Solution Approach 2:
The patent introduces identifier fields that link parent and child documents across separate indexes. These identifier intermediaries enable the system to correlate and query related documents from different indexes by matching identifiers, thus maintaining index separation while enabling cross-index queries
2Device complexity
If parent and child documents are indexed separately, then index maintenance is simpler, but search operations require multiple queries and more resources
Solution Approach 1:
The patent performs preliminary data preparation by embedding child document information in parent documents during the indexing phase. This preliminary action consolidates related data beforehand, so that subsequent search operations can retrieve both parent and child data in a single query without requiring multiple separate index accesses
Solution Approach 2:
The patent merges parent and child document data into a unified index structure where parent documents contain embedded child document information. This merging allows the system to maintain relatively simple index structures while enabling efficient single-query retrieval of related documents, improving search productivity
Data Source
AI summary
Systems and methods are described for indexing related documents to provide more robust search functionality within an index. In one aspect, a first document, such as a comment or status update associated with a second document or ticket item, may be obtained. As a result of the first document including first data that is linked to second data of the second document, an instance of the second document may be modified to include at least a portion of the first data. An index, which indexes a corpus of documents, such as ticket items, may be updated to include the modified second document, such that the first data is associated with the modified second document in the index. Responsive to a query for the first data in the index, the modified second document and any included first data may be returned.