Data-SKUP 2024

Flyspotting er ikke lenger bare for entusiastene og nørdene. Informasjon om fly- og helikopterbevegelser har blitt et uvurderlig verktøy med stort potensiale, ikke minst for grave- og breakingjournalister. I dette foredraget vil du få innsikt i flere journalistiske prosjekter der flytrafikkdata har spilt en avgjørende rolle. Du vil også lære hvordan du enkelt kan sette opp en egen radarstasjon hjemme eller på kontoret, og hvordan du kan få gratis tilgang til premiumtjenester som Flightradar24, Radarbox og FlightAware ved å dele dine data. Vi går også gjennom hvordan du kan bruke sanntidsdata for å bli varslet når bestemte fartøy observeres i nærområdet ditt, og hvordan du kan analysere millioner av historiske datapunkter og gjøre spennende funn.

Norge har et stort antall lokalaviser, men journalistikken er under press, og økonomien er utfordrende. Kan samarbeid mellom redaksjoner øke kvaliteten på journalistikken? 
På Samarbeidsdesken arbeider fire personer arbeider på vegne av 122 redaksjoner. De lager undersøkende journalistikk, som de deler med alle redaksjoner, som deretter tilpasses til lokale forhold.
Øivind Skjervheim fra Samarbeidsdesken kommer til Dataskup for å fortelle om ideen og prosjektene de har arbeidet med så langt.

Earth Index is an innovative tool developed by the nonprofit Earth Genome that uses satellite imagery and machine learning to make environmental data more accessible and actionable for journalists and researchers.
The tool features a map-based interface, allowing users to search for specific environmental features with ease. Its AI-powered analysis processes vast amounts of Earth observation data, while data compression technology reduces the volume by up to 10,000 times, retaining only essential information. Earth Index also identifies patterns, such as mining operations, airports, or farms, making it a powerful tool for spotting environmental trends.
One notable application of Earth Index is the Amazon Mining Watch, where it has been used to detect illegal gold mining in the Amazon, supporting investigations and even legal actions against violators. It also enables real-time environmental monitoring and empowers users to create impactful datasets, putting the power of data into the hands of those who can drive change.

The GAIJ project is a 15-month pilot initiative aimed at developing an open-source AI tool to help investigative journalists and the public identify financial fraud. Led by the Artificial Intelligence Journalism Resource Center (AIJRC) and the Norwegian University of Life Sciences, the project utilizes Large Language Models (LLMs) and graph analysis to detect suspicious interactions in financial records, such as tax declarations and financial transactions.
By combining AI and machine learning, GAIJ will create a "suspicion graph" that flags irregular financial transactions between entities, making it easier for journalists to expose corruption and non-sustainable practices. This project is particularly relevant to the UN’s Sustainable Development Goals, as it aims to combat illicit financial flows and promote transparency. GAIJ will work closely with investigative journalists, especially from Norwegian media, to ensure the tool is practical and impactful.

This talk delves into the principles behind creating intuitive tools that allow journalists to explore large datasets effectively, with a focus on making AI accessible and understandable. By embedding AI functionalities within user-friendly interfaces, journalists can leverage powerful data analysis and pattern recognition without requiring deep technical expertise. We will showcase use cases to demonstrate how these design principles bridge the gap between journalism and advanced AI, enabling reporters to uncover insights, and generate data-driven stories more efficiently. Particular relevance will be given to privacy and ethical consideration.