Advanced search
Start date
Betweenand

Automatic adaptation of annotations in the semantic web

Grant number: 19/14582-8
Support type:Scholarships in Brazil - Master
Effective date (Start): August 01, 2019
Effective date (End): July 31, 2021
Field of knowledge:Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
Principal Investigator:Julio Cesar dos Reis
Grantee:Enio de Jesus Pontes Monteiro
Home Institution: Instituto de Computação (IC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
Associated research grant:17/02325-5 - EvOLoD: linked data evolution on the Semantic Web, AP.JP

Abstract

The access and use of metadata semantically defined can benefit many types of computational tasks. Data integration between different systems and search engines can exploit the semantic definitions associated with documents and other Web resources. The generation of annotations from the growing number of interconnected RDF repositories makes available the required metadata to enable more interesting applications on the Web. For example, semiautomatic tools recognize named entities as people and places in Web documents. However, the semantic definition of these entities tends to change in their bases and implies in a scenario highly challenging for the Semantic Web. Therefore, generated annotations need to be updated due to the evolution of knowledge bases. This master's research aims to investigate adaptation techniques of annotations defined from structured data interconnected in the Web. The adaptation approach will be based on types of modifications detected in the evolution of knowledge bases. To this end, experimental analyzes with real data will be carried out to deeply understand the relationship between change operations on RDF bases and the way that annotations change over time. These analyzes should inform the design and formalization of adaptation operations that when applied to annotations change their elements and update their state. The decision process will involve the use of heuristics and rules that formalize the conditions necessary to apply the adaptation operations. We will develop algorithms and software prototypes implementing the methods of adapting semantic annotations in Linked Data. The research includes an experimental evaluation of the concept proof developed in computational tools. The evaluation methodology covers relevant scenarios to test the quality of the adapted annotations. The experimental validation should involve several configurations and success measures. Objective metrics such as precision and coverage will indicate the efficiency of the results obtained according to a set of gold standards. The main expected contribution is an automatic method of annotation adaptation based on types of modifications characterized and identified in the evolution of knowledge bases in RDF. Results obtained will allow to maintain the evolution consistency of semantic annotations affected by the evolution of knowledge bases. (AU)