Service Request Analyzer Extracting Knowledge from Unstructured Text
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies face challenges in automating the extraction and reuse of knowledge from unstructured data, particularly in service centers, where vast amounts of unstructured text contain valuable information but are difficult to process due to duplicates, irrelevant content, and diverse user needs, leading to inefficiencies in service delivery.
Innovation Solution
The implementation of a system that includes a Service Request Analyzer and Recommender (SRAR) with a hierarchical classifier and feature generator, utilizing domain knowledge and human expertise to extract relevant knowledge nuggets (KN) and process nuggets (KP) from unstructured text, categorizing and recommending relevant service requests, and utilizing a Knowledge Process Miner to refine extracted knowledge.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual knowledge extraction from unstructured data is performed, then knowledge accuracy is maintained, but productivity is low and time consumption is high
Solution Approach 1:
The patent introduces an intermediary system comprising a hierarchical classifier and feature generator that acts as a mediator between unstructured service request data and knowledge extraction. The system pre-processes and structures unstructured data into standardized formats, enabling automated knowledge extraction while maintaining accuracy. This intermediary layer resolves the contradiction by automating time-consuming manual tasks while preserving knowledge quality through structured transformation.
Solution Approach 2:
The system performs preliminary actions by pre-processing unstructured service request data before knowledge extraction. The hierarchical classifier and feature generator prepare data in advance by categorizing and structuring it, so that when knowledge extraction occurs, it can be done rapidly and accurately. This preliminary structuring eliminates the need for time-consuming manual processing during actual knowledge extraction.
2Productivity
If automated processing of unstructured data is implemented, then productivity is improved, but measurement precision and reliability of knowledge extraction deteriorate
Solution Approach 1:
The patent applies segmentation by dividing the automated processing system into hierarchical levels: a first-level classifier for broad categorization and a second-level classifier for detailed classification. This segmented approach allows automated processing to maintain precision by handling different aspects of data at appropriate levels of granularity, ensuring accurate knowledge extraction while improving productivity through systematic automation.
Solution Approach 2:
The system changes parameters by transforming unstructured data into structured formats with specific parameters and features. The feature generator extracts and standardizes key parameters from unstructured text, enabling automated processing to achieve both high productivity and measurement precision by working with well-defined data parameters rather than raw unstructured content.
3Loss of information
If comprehensive knowledge extraction from all data sources is performed, then knowledge completeness is improved, but device complexity and processing requirements increase
Solution Approach 1:
The patent applies the extraction principle by selectively removing and focusing on relevant knowledge elements from unstructured data sources. The hierarchical classifier and feature generator extract only the most important and relevant features and categories, rather than processing all data comprehensively. This selective extraction maintains knowledge completeness for critical information while reducing system complexity by ignoring irrelevant data.
Solution Approach 2:
The system performs partial action by focusing knowledge extraction on the most critical and relevant portions of unstructured data. Rather than attempting to extract and process all possible information comprehensively, the hierarchical classifier targets specific high-value knowledge elements, achieving sufficient knowledge completeness for service delivery while avoiding the excessive complexity of comprehensive processing.
Data Source
AI summary
Methods and systems for knowledge extraction that involve providing analytics and blending the analytics with analysis of one or more knowledge processes are provided. Knowledge extraction may be based on combining analytic approaches, such as statistical and machine learning approaches. Unstructured data, such as numerical, geo-spatial, text, speech, image, video, data, and music, may be used as input for these processes. The methods and systems may convert this unstructured data into a structured knowledge that has some specific utility to its user. Some embodiments may involve service requests delivery, information and knowledge extraction, information and knowledge retrieval, media mining, marketing, and other uses. Different granularity levels of knowledge and information extraction may be provided. This differentiation may be used for monetization of the service.