Academic research use — rate limit increase and historical retrieval advice

[Amirhossein Hanifirad]

Hello Open Platform team,

I'm an MSc Advanced Computer Science student at the University of Kent, writing my dissertation on whether geopolitical news sentiment improves machine-learning prediction of next-day Brent crude oil returns. The Guardian Content API is central to it: I'm building a daily news-sentiment feature from articles across 1999–2026, which an LLM then converts into a structured daily signal.

I have two questions I'd be very grateful for help with.

1. Rate limit. During bulk historical retrieval my developer key hits the daily limit. As this is non-commercial academic work, would it be possible to raise the daily limit on my key, even temporarily, so I can complete the historical fetch?

2. Best-practice for a relevant subset. Rather than pulling everything, I want the articles genuinely relevant to oil and geopolitics (oil markets, OPEC, major producer countries, shipping chokepoints, sanctions, conflict). I've been combining section and tag filters with the q parameter, but I'd value your guidance: is there a recommended way to retrieve a focused, relevant historical subset reliably — for example a preferred combination of tags, sections, and query terms, or a bulk/export route for academic use — that avoids the 10,000-results-per-query pagination limit?

The work is entirely non-commercial and I'll acknowledge the Guardian Open Platform in my dissertation. My developer key is registered to this email address.

Thank you very much — the API is a huge help to this research.

With thanks,
Amir
MSc Advanced Computer Science, University of Kent