Magazine article AI Magazine

Information Integration on the Web

Magazine article AI Magazine

Information Integration on the Web

Article excerpt

Information-integration techniques enable interaction between users and data sources through a centralized access point and uniform query interfaces that give users the illusion of querying a homogeneous system. Most integration solutions have assumed structured sources with the heterogeneity being introduced by the variety in source schemas and data models. But a large portion of the web consists of pages that contain information presented as unstructured text, such as, for example, blogs, wikis, reviews, and so on. Therefore, integration systems that can match entities or objects across both structured and unstructured sources are the need of the hour. Recent research on web object extraction, record linkage, and named entity recognition has generated some initial solutions; however, many challenges remain in developing such a system.

This workshop, sixth in the IIWeb series, was proposed to bring together researchers looking to effectively integrate information from both structured and unstructured sources of information. The anticipated outcome of the workshop was to assess the state of the art in the area, as well as to identify critical next steps to pursue in this topic.

Accordingly we asked submissions in topics such as automatic wrapper induction, schema matching, web entity extraction and search, record linkage and object consolidation, database and information retrieval (IR) integration, applications, and experiences. The call attracted 24 submissions out of which 12 were selected for full presentation and 6 for short presentation. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.