Detecting Non-Obvious Hydrocarbon Plays from Unstructured Text
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting Hydrocarbon Plays from unstructured text fail to identify non-obvious combinations of Hydrocarbon Play Elements, ignore concepts without geological age, aggregate concepts independently, and lack differentiation between well-known and speculative statements, leading to missed potential Hydrocarbon Play detection.
Innovation Solution
A data processing method and system that detects Hydrocarbon Play Elements in sentences, preserves their relative sequence, and computes a rank score for their non-obvious nature using machine learning, natural language processing, and lexicons to identify potential Hydrocarbon Plays within and across documents.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If current methods aggregate concepts independently without preserving sequence, then processing is simpler, but detection of combinatorial sequences suggesting potential Hydrocarbon Plays is lost
Solution Approach 1:
The patent segments the text processing into distinct stages: first extracting individual Hydrocarbon Play Elements with their contexts, then separately analyzing their sequences and combinations. This segmentation allows independent processing of concepts while preserving their original sequence information for subsequent combinatorial analysis, resolving the contradiction between processing simplicity and information retention.
Solution Approach 2:
The patent adds a temporal/sequence dimension to the analysis by preserving the order of Hydrocarbon Play Elements as they appear in text. Instead of treating concepts as a simple set, the system maintains their sequential arrangement, enabling detection of combinatorial patterns that depend on specific ordering, thus preventing information loss while maintaining processing efficiency.
2Ease of manufacture
If all concepts are treated equally without differentiation, then processing is more uniform, but differentiation between well-known and speculative statements is lost
Solution Approach 1:
The patent applies local quality by differentiating the treatment of individual Hydrocarbon Play Elements based on their specific contexts. Each element is analyzed for its non-obvious nature using domain-specific clues and machine learning, allowing the system to maintain uniform processing workflows while applying context-sensitive differentiation to identify speculative versus well-known statements.
3Measurement precision
If concepts with geological age are prioritized, then geological accuracy is improved, but concepts without geological age are ignored
Solution Approach 1:
The patent applies partial action by selectively applying geological age filtering only when relevant to the analysis. The system extracts and processes all Hydrocarbon Play Elements regardless of whether geological age is present, then applies age-based filtering only in contexts where it enhances detection accuracy. This ensures concepts without geological age are not lost, while still utilizing age information where available for improved precision.
4Measurement precision
If manual LBD methods are used, then detection of non-obvious relationships is improved, but processing of millions of documents is not feasible
Solution Approach 1:
The patent introduces machine learning models and domain-specific lexicons as intermediaries between manual analysis and automated processing. These intermediaries capture the nuanced detection capabilities of manual LBD methods while enabling scalable processing of large document volumes. The system uses trained models to identify non-obvious relationships automatically, bridging the gap between precision and productivity.
5Reliability
If deductive techniques with a priori lexicons are used, then associations between elements are derived systematically, but inductive techniques for deriving concepts directly from text are not utilized
Solution Approach 1:
The patent merges deductive and inductive techniques into a unified system. It uses a priori lexicons and taxonomies for systematic association derivation while simultaneously applying topic modeling and machine learning to discover new concepts and associations directly from text patterns. This combination allows the system to benefit from both the reliability of systematic methods and the adaptability of data-driven discovery.
Data Source
AI summary
The invention is a data processing method and system for suggesting non-obvious potential Hydrocarbon Plays from unstructured text. The data processing system detects clues for Hydrocarbon Play Elements in sentences, matches combinatorial patterns (DNA inspired) across sentences and documents to output potential Hydrocarbon Plays. The system also computes a rank for the detected Hydrocarbon Play's non-obvious nature. In particular, the data processing system operates on any digital unstructured text derived from academic literature, company reports, web pages and other sources. Detected Hydrocarbon Plays can be used to stimulate ideation and learning events for geoscientists in the oil and gas exploration industry.