Generative ML Models for Genealogical Summary Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large genealogy databases face challenges in identifying and linking related datasets due to discrepancies in historical records and human accounts, making it difficult for users to understand their family history and connect ancestral data with historical context.
Innovation Solution
A machine-learning language model generates shareable genealogical summaries by retrieving and processing genealogical records, converting hierarchical data structures into natural language descriptions, and inputting embeddings into a generative model to produce summaries that describe relationships and life events of ancestors, including contextual information and historical details.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If users manually compare large numbers of datasets to identify relatives, then identification accuracy improves, but computational feasibility deteriorates
Solution Approach 1:
The patent introduces an intermediary representation layer (embeddings, graphs, or summaries) that mediates between raw datasets and the identification task. Instead of directly comparing all datasets, the system converts them into intermediate representations that capture essential relationships, enabling efficient relative identification without manual comparison of all data.
Solution Approach 2:
The patent extracts key relationship information from the full datasets by identifying and isolating specific connection patterns (e.g., shared ancestors, relationship paths). This extraction allows the system to focus computational resources on analyzing only the relevant relationship structures rather than processing entire datasets manually.
2Quantity of substance
If the system processes all genealogical records to ensure comprehensive coverage, then data completeness improves, but processing time deteriorates
Solution Approach 1:
The patent segments the genealogical data into manageable units (individual records, family clusters, or relationship paths) and processes them in parallel or hierarchical fashion. This segmentation enables comprehensive data coverage while reducing processing time by handling small units efficiently rather than processing the entire dataset as one large unit.
Solution Approach 2:
The patent performs preliminary processing actions such as pre-computing relationship paths, pre-indexing data structures, or pre-generating summaries before the main analysis task. This preliminary action reduces the computational burden during the main processing phase, enabling comprehensive data coverage with reduced overall processing time.
3Measurement precision
If the system links all possible datasets to maximize relationship identification, then identification accuracy improves, but device complexity deteriorates
Solution Approach 1:
The patent introduces intermediate data structures (such as relationship graphs or embedding spaces) that simplify the linking process. These intermediaries organize and pre-process the connections between datasets, reducing the complexity of the final data structure while maintaining comprehensive relationship identification accuracy.
Solution Approach 2:
The patent merges multiple data representation formats into a unified structure that can handle diverse genealogical records efficiently. By combining different data types (textual records, relationship paths, and metadata) into a unified representation system, the patent reduces overall system complexity while maintaining comprehensive relationship linking capabilities.
Data Source
AI summary
Disclosed herein are methods, systems, and non-transitory computer readable mediums for generating a shareable genealogical summary for a target individual. An example method includes receiving a request from a user to generate a shareable genealogical summary about a target user. The method generates the shareable genealogical summary comprising a genealogical history of the target user. The method provides genealogical information for the target user to a machine-learning language model. The genealogical information includes a family tree. The method receives a response generated by executing the machine-learning language model from a model serving system. The method provides the shareable genealogical summary for display to the user.