17534489. SEMANTIC ANNOTATION FOR TABULAR DATA simplified abstract (International Business Machines Corporation)
Contents
SEMANTIC ANNOTATION FOR TABULAR DATA
Organization Name
International Business Machines Corporation
Inventor(s)
Udayan Khurana of White Plains NY (US)
Sainyam Galhotra of Somerville MA (US)
SEMANTIC ANNOTATION FOR TABULAR DATA - A simplified explanation of the abstract
This abstract first appeared for US patent application 17534489 titled 'SEMANTIC ANNOTATION FOR TABULAR DATA
Simplified Explanation
The patent application describes an approach to mapping columns of data to semantic concepts using maximum likelihood estimation and publicly available structured data.
- The approach includes specialized estimation methods for different types of data (categorical, numeric, alphanumeric) within a common framework of likelihood estimation.
- Indexes are used to support quick estimation computations for different types of data.
- The approach allows for the utilization of semantic context without a significant increase in mapping runtime or resource utilization.
Potential Applications
This technology has potential applications in various fields, including:
- Data analysis and data mining
- Natural language processing and text mining
- Information retrieval and search engines
- Machine learning and artificial intelligence
Problems Solved
The approach solves the following problems:
- Efficiently mapping columns of data to semantic concepts
- Handling different types of data (categorical, numeric, alphanumeric) in a unified framework
- Utilizing semantic context without a significant increase in runtime or resource utilization
Benefits
The benefits of this technology include:
- Improved accuracy and efficiency in mapping data to semantic concepts
- Ability to handle different types of data within a common framework
- Utilization of semantic context without a significant impact on performance or resource usage
Original Abstract Submitted
An approach to column to semantic concept mapping using joint estimation through piecewise maximum likelihood estimation and utilizing large openly available structured data may be provided. The approach may include a special estimation methods for categorical, numeric, and alphanumeric/symbolic data, while unifying the overarching estimation with a common framework of likelihood estimation. The approach may also include indexes to support quick estimation computations for numeric, categorical, and mixed type data. Additionally, the approach may include semantic context utilization without a polynomial increase in mapping runtime or resource utilization.