Extracting assumptions from missing data
From Semantic Portal Wiki
{{#vardefine:category|Publication}}{{#vardefine:templatename|i.publication}}{{#vardefine:package|smwbp_instance_templates}}
| Edit |
Reference: {{#vardefine:pagename|extracting assumptions from missing data }}
- [[]]
bibtex
{{#vardefine:pagename|Extracting assumptions from missing data }}{{#vardefine:key| }}
abstract: Information integration is the task of aggregating data from multiple heterogeneous data sources. The understandings of context knowledge of data sources are often the keys to challenging problems in information integration such as handling missing and inconsistent data. Context logic provides a uni-fied framework for the modeling of data sources; nevertheless, the acquisition of large amounts of context knowledge is difficult. In this paper, we study the importance of a special type of context knowledge, namely assumption knowl-edge. Assumption knowledge refers to a set of implicit rules about assump-tions on which a data source is based. We develop a decision tree classifier to extract assumption knowledge from missing data and formalize the knowledge in context logic. Finally, we build an information aggregator with assumption knowledge reasoning, which is capable of explaining incomplete data aggre-gated from heterogeneous sources.
download:
- paper:
- slides:
| Abstract | Information integration is the task of agg … Information integration is the task of aggregating data from multiple heterogeneous data sources. The understandings of context knowledge of data sources are often the keys to challenging problems in information integration such as handling missing and inconsistent data. Context logic provides a uni-fied framework for the modeling of data sources; nevertheless, the acquisition of large amounts of context knowledge is difficult. In this paper, we study the importance of a special type of context knowledge, namely assumption knowl-edge. Assumption knowledge refers to a set of implicit rules about assump-tions on which a data source is based. We develop a decision tree classifier to extract assumption knowledge from missing data and formalize the knowledge in context logic. Finally, we build an information aggregator with assumption knowledge reasoning, which is capable of explaining incomplete data aggre-gated from heterogeneous sources. ta aggre-gated from heterogeneous sources. |
| Address | Paris + |
| Author | Honglei Zeng +, and Richard Fikes + |
| Bibtype | inproceedings + |
| Booktitle | Context representation and reasoning 2005, proceedings of the first international workshop + |
| Key | KSL-05-07 + |
| Month | July + |
| Tag | Computer science + |
| Title | Extracting Assumptions from Missing Data + |
| Tr id | KSL-05-07 + |
| Year | 2005 + |
