End-to-end solution for linked open data query logs analytics

TL;DR

The paper addresses the challenge of extracting trustworthy insights from linked open data query-logs by proposing an end-to-end, four-layer architecture that emphasizes trust throughout. It details a Layer 0 raw-log store, a Layer 1 preparation and curation process (including log extraction, profiling, and trust annotation), a Layer 2 storage, and a Layer 3 analytics layer that derives multidimensional patterns and a data warehouse via LogLinc and semantic similarity. Experiments on Scholarly data and DBpedia logs demonstrate substantial trust mejoras after curation (e.g., 81% and 65% trusted queries, respectively) and show the feasibility of generating a decision-support data warehouse from LOD logs. The work highlights practical significance for enterprises and researchers seeking reliable analytics from LOD query-logs and sets the stage for a trust-centered tooling ecosystem. Future directions include developing a dedicated trust-based tool and extending analytics to broader business contexts.

Abstract

Important advances in pillar domains are derived from exploiting query-logs which represents users interest and preferences. Deep understanding of users provides useful knowledge which can influence strongly decision-making. In this work, we want to extract valuable information from Linked Open Data (LOD) query-logs. LOD logs have experienced significant growth due to the large exploitation of LOD datasets. However, exploiting these logs is a difficult task because of their complex structure. Moreover, these logs suffer from many risks related to their Quality and Provenance, impacting their trust. To tackle these issues, we start by clearly defining the ecosystem of LOD query-logs. Then, we provide an end-to-end solution to exploit these logs. At the end, real LOD logs are used and a set of experiments are conducted to validate the proposed solution.

Figures (1)

• Figure 1: End-to-End Architecture for LOD logs analytics