Technology-Driven Data Journalism: Transparency and Accuracy Mechanisms for Innovation and Regulation of the Media in Albania Aligned with EU Standards

by Anxhela FERHATAJ

Abstract
Purpose: This study analyzes three web applications for interrogating information flows and narratives. It examines an explanatory web app derived from the book “From Isolation to Integration: Media Discourse on Albania’s EU Accession: Monitoring National TV Narratives and Public Perceptions”, Tools for Innovation Monitoring Analytics as a policy-trend and weak-signals dashboard, and OSoMeNet as a diffusion-mapping tool. Design/methodology/approach: Comparative, multi-layer design across three applications: an explanatory web app derived from the book “From Isolation to Integration: Media Discourse on Albania’s EU Accession: Monitoring National TV Narratives and Public Perceptions”, TimAnalytics for research and policy signals (2015–2025), and OSoMeNet for recent online diffusion. TimAnalytics outputs are deduplicated and aggregated by year and document type. OSoMeNet interaction networks (replies, mentions, reposts) yield giant-component share, modularity, betweenness of bridging hubs, cascade size and depth, and time. Retrieval dates, queries, and export hashes are archived for reproducibility.

Findings: TimAnalytics shows sustained growth in fake-news detection after 2018 and a second wave in 2022–2024, a 2025 tilt toward large-language-model and multimodal approaches, and a geography concentrated outside the Western Balkans. OSoMeNet reveals a sparse, modular network for the most recent three months (1,464 nodes; 1,511 edges; average degree 2.06; density 0.14%), with mentions outweighing replies and bridging hubs driving cross-community reach. The book-derived app documents outlet differences in agenda, frames, and tone and flags claim-dense TV segments that often precede online spread. Together, the tools increase transparency through source-linked visuals and audit trails, improve detection accuracy when paired with local evaluation sets and error analysis by outlet and claim type, and enable timelier corrections by monitoring core communities and routing through hubs.

Originality: An integrated assessment of three applications (TV framing, research-policy signals, online diffusion) produced convergent, reproducible metrics supporting transparent pipelines, accurate detection, and timely corrections in newsroom practice.

Keywords: data journalism, web applications, misinformation detection, network diffusion analysis, transparency, correction timeliness