Dataset

machine-readable-news-scores

Automated news feed with sentiment ratings, delivered in machine-readable form via the LSEG (formerly Refinitiv) distribution network.

Visit provider

Dataset card

LSEG News Analytics' machine-readable-news-scores dataset provides precomputed sentiment and relevance scores derived from news content for use in financial analytics workflows. It is distributed under a commercial license.

What this dataset contains

The dataset supplies scored outputs produced by LSEG News Analytics' news-processing pipeline, covering items tagged with machine-readable-news, sentiment, and refinitiv themes. The source is the same engine that powers Refinitiv news analytics, making the scores compatible with downstream financial-market applications. Specific column definitions, row counts, and update cadence are not documented in the available metadata.

Structure and fields

The official schema and field names are not disclosed; the facts list columns as (unknown) and no masked sample rows are available. Because the underlying service is Refinitiv's news scoring feed, typical deployments expose identifiers, timestamps, headline text, and per-item sentiment and relevance scores, but these field names cannot be confirmed from the supplied metadata.

How analysts use it

Machine-readable news scores are typically ingested into systematic trading and risk models to convert unstructured press coverage into quantitative signals. The available facts do not specify coverage, history, or asset-class scope, so concrete fund-use cases cannot be claimed with confidence. The commercial license implies the data is purchased through a licensed feed rather than downloaded directly.

Documentation & references 2