A process framework for inducing and explaining Datalog theories
| Autor: | Mark Gromowski, Michael Siebers, Ute Schmid |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Statistics and Probability
Process (engineering) Computer science Applied Mathematics Retraining 02 engineering and technology Data science Computer Science Applications Datalog Inductive logic programming Order (exchange) 020204 information systems 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Everyday life Rule of inference computer Natural language computer.programming_language |
| Zdroj: | Advances in Data Analysis and Classification. 14:821-835 |
| ISSN: | 1862-5355 1862-5347 |
| DOI: | 10.1007/s11634-020-00422-7 |
| Popis: | With the increasing prevalence of Machine Learning in everyday life, a growing number of people will be provided with Machine-Learned assessments on a regular basis. We believe that human users interacting with systems based on Machine-Learned classifiers will demand and profit from the systems’ decisions being explained in an approachable and comprehensive way. We developed a general process framework for logic-rule-based classifiers facilitating mutual exchange between system and user. The framework constitutes a guideline for how a system can apply Inductive Logic Programming in order to provide comprehensive explanations for classification choices and empowering users to evaluate and correct the system’s decisions. It also includes users’ corrections being integrated into the system’s core logic rules via retraining in order to increase the overall performance of the human-computer system. The framework suggests various forms of explanations—like natural language argumentations, near misses emphasizing unique characteristics, or image annotations—to be integrated into the system. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink