About the Company
World Justice Project (WJP) is an independent, multidisciplinary non-profit organization headquartered in Washington D.C., with additional offices in Mexico City and Singapore. The organization is dedicated to advancing the rule of law worldwide through creating knowledge, building awareness, and stimulating action. WJP focuses on promoting fundamental freedoms, justice, and governance globally by conducting extensive research and data collection, primarily through perception surveys of the public and experts. We had the pleasure to speak to Carlos Toruno, a senior data analyst at WJP, who shared his journey with NewsCatcher.
Identifying the Problem
Our organization faced a significant challenge in validating survey data collected globally to assess democracy and governance. Currently, in the European Union, we collect public opinion polls from the 27 member countries, gathering between 500 and 8000 surveys per country, depending on its size. Additionally, we collect data from experts, such as lawyers and academics, on topics like governance, civil and commercial law, as well as criminal justice.
To validate this public and expert perception-based data, we needed to assess and understand the context and potential impact of recent political events in these countries, making news articles a critical input. Usually, this process involves hiring regional experts to provide qualitative cross-check reports on an annual basis. These experts used a combination of news articles and reports from international organizations, mainly in English.
“To enhance and make this process more efficient, we were looking for a solution that would give us access to local news articles and go beyond sources in just English.”
That’s when we started trying a news API to provide us with the inputs to set up a political events tracker through AI. Our goal was to use large language models to classify and detect political events as narrated by local media.
Adopting the Solution
Initially, we used a different news API but faced several issues, such as receiving incomplete content from articles, the exclusion of many targeted newspapers from the API's database, and limited results when using specific keywords related to democracy, freedoms, and justice.
“After trying another API for two months with limited success, we switched to NewsCatcher. The results were remarkable.”
From our target list of 220 newspapers in the 27 EU member countries, we managed to access 200 newspapers with NewsCatcher, compared to only 100-120 with the previous API. The number of relevant news articles increased significantly, and issues like empty content were virtually non-existent with NewsCatcher.
This solution enabled us to improve the data validation process, ensuring more comprehensive and accurate analysis by incorporating a wider range of local news sources across different languages.
Implementation
The detailed information provided on NewsCatcher's website has been instrumental in guiding our decision-making process. The transition from the other news API to NewsCatcher was smooth, and we immediately noticed impressive results.
Upon implementing NewsCatcher, we quickly observed a substantial increase in the amount of high-quality data, allowing us to proceed with the costly and complex AI classification process.
“The diverse range of sources and the high quality of content provided a robust foundation for our data validation, ensuring comprehensive and reliable analysis.”
We targeted a list of 220 influential newspapers across the 27 EU countries, selected based on diversity in editorial lines, historical influence, editorial influence, and daily prints. We used specific keywords to filter news articles and used large language models (LLMs) to classify these articles further.
Conclusion
The adoption of NewsCatcher has enhanced WJP’s overall operations, providing a reliable and efficient solution for data validation and helping them better fulfil WJP’s mission of promoting the rule of law worldwide.
This success has motivated Carlos and his team to explore further applications of NewsCatcher in other research areas of the organization. For example, they plan to assess media coverage of the upcoming US elections and analyze related news data using their NLP techniques. Additionally, they are exploring the use of NewsCatcher to evaluate environmental governance by targeting news articles about environmental scandals and regulatory issues.
One last note from Carlos…
Initially, they were hesitant to schedule a call - they just wanted a quick solution.
“But jumping on a call with NewsCatcher proved to be incredibly valuable. The personalized service and tailored solutions made a significant difference, proving that the initial effort to engage with them was well worth it.”