We are in the process of creating a market sentiment dashboard per currency pair and on certain commodities using Python programming. This program runs on real-time FX market news from Refinitiv. We would like the filter the news by those having a short-term and long-term impact on that currency pair/commodity. How can this be done please?
Since you posted your question on the forum dedicated to Eikon Data APIs, I assume you're using Eikon. If you're using another Refinitiv product, please specify which Refinitiv product or API you're utilizing.
Eikon Data APIs provide capability to retrieve news headlines and stories. However, this API does not have the capability to assess what kind of impact news headline or story has on the market. If you're considering developing your own analytics to determine the likely market impact of a news story, there's a couple of videos that discuss NLP and basic sentiment analysis on news stories retrieved using Eikon Data APIs under the Tutorials section for Eikon Data APIs on Refinitiv Developer Portal. You may wan to start by watching these videos.
If you're looking for Refinitiv to provide sentiment analysis on news headlines, I suggest you take a look at Refinitiv Machine Readable News Proposition, which allows you to consume Refinitiv news in a custom enterprise application and includes market sentiment indicators for news. To the best of my knowledge it does not provide the assessment of short term vs. long term market impact. I've never heard of such capability available anywhere. But I'm not an expert in NLP. It may very well be that there's some NLP software available somewhere that is capable of assessing short term vs. long term market impact of news.