6533b851fe1ef96bd12a8aab

RESEARCH PRODUCT

DYNAMIC SEMANTIC USER PROFILING FROM IMPLICIT WEB NAVIGATION DATA

Anett HoppeAna RoxinChristophe Nicolle

subject

JEL classification: M37 Advertising; L86 Information and Internet Services Computer Software; D80 General (Information Knowledge Uncertainty)Web Analysis[INFO.INFO-CL] Computer Science [cs]/Computation and Language [cs.CL]Rule-based reasoningOntologiesUser Profiling[ INFO.INFO-CL ] Computer Science [cs]/Computation and Language [cs.CL][INFO.INFO-CL]Computer Science [cs]/Computation and Language [cs.CL]Semantic Web

description

International audience; On the Web, pages are often dynamically generated and allow publishers to individually adapt contents to each viewer. Underlying systems must correctly understand the user's context - crucial especially in the case of online advertisement placement. The article at hand describes our proposition of a novel profiling system, adapted to the special needs of digital advertising. Based on Semantic Web Technologies, the MindMinings system relies on an ontology to enable thorough understanding of each user's context and needs. The underlying ontology structure also provides enhanced interoperability with semantically annotated knowledge resources, notably vocabularies from the Linked Open Data cloud. All components in the context of this project have been defined in close collaboration with an industrial partner from the digital advertising domain. Therefore, developments made for the system have been verified to be adapted to an application in an industrial context. In this article, we describe system's architecture, issues encountered during its development and envisaged solutions.

yearjournalcountryeditionlanguage
2014-05-15
https://hal.archives-ouvertes.fr/hal-00978436