Aggregate article counts

Search query using Lucene-like syntax: supports boolean operators (AND, OR, NOT, and &&, ||, ! as aliases), exact phrase matching with double quotes, grouping with parentheses, and wildcards. Cannot be empty.

Which fields to search the query against. Up to 2 values from: title, content, title_content, summary (summary requires an NLP plan). NLP/translation plans additionally allow searching translated fields: title_translated, content_translated, summary_translated, title_content_translated.

Filter by how precisely the published date is known: 'full' (exact date and time), 'date' (day known, time unknown), or 'timezone unknown' (date and time known, but not the timezone).

When true, the date range and sorting apply to the date NewsCatcher parsed/ingested the article instead of its published date.

Sort order for results: 'relevancy' (default, best text match first), 'date' (newest first), '-date' (oldest first), or 'rank' (most authoritative source first).

Minimum source rank (lower rank = more authoritative/popular source). Integer between 1 and 999999.

Maximum source rank (lower rank = more authoritative/popular source). Integer between 1 and 999999.

Minimum article word count. Must be non-negative, and less than word_count_max if both are set.

Maximum article word count. Must be non-negative, and greater than word_count_min if both are set.

Page number of results to return, starting at 1.

Number of articles per page. Integer between 1 and 1000.

Which article field to base clustering similarity on: 'content', 'title', or 'summary'.

Similarity threshold for clustering, greater than 0 and up to 1 inclusive. Higher values require greater similarity to group articles into the same cluster.

For NLP-enabled plans, controls whether the nlp object is included in the response for each article (excluded by default). Basic plan always includes nlp.theme regardless of this setting.

When true, only returns articles that have NLP enrichment data. Requires an NLP plan.

Filter by article theme. One of: Business, Politics, Entertainment, Sports, Finance, Crime, Tech, General, Health, Economics, Science, Automotive, Weather, Travel, Lifestyle, Financial Crime. Comma-separated values are OR'd together; values joined with ' AND ' within one comma-separated group must all match the same article (e.g. 'Tech AND Finance,Sports' matches Tech-and-Finance articles, or Sports articles).

Exclude articles with any of the given themes. Same allowed values as theme, but does not support ' AND ' grouping - always excludes on any match.

Filter articles that mention the given organization name(s) in NLP-extracted named entities. Requires an NLP plan.

Filter articles that mention the given person name(s) in NLP-extracted named entities. Requires an NLP plan.

Filter articles that mention the given location name(s) in NLP-extracted named entities. Requires an NLP plan.

Filter articles that mention the given miscellaneous entity name(s) in NLP-extracted named entities. Requires an NLP plan.

Minimum sentiment score for the article title, from -1 (most negative) to 1 (most positive), as determined by NLP analysis. Requires an NLP plan.

Maximum sentiment score for the article title, from -1 (most negative) to 1 (most positive), as determined by NLP analysis. Requires an NLP plan.

Minimum sentiment score for the article content, from -1 (most negative) to 1 (most positive), as determined by NLP analysis. Requires an NLP plan.

Maximum sentiment score for the article content, from -1 (most negative) to 1 (most positive), as determined by NLP analysis. Requires an NLP plan.

Deprecated alias for aggregate_by - only one of the two may be set; whichever is set is copied to the other.

Time interval to aggregate article counts by: 'day', 'hour', 'week', or 'month'.

Always 'ok' for a successful aggregation request.

Total number of articles matching the query filters, across the full date range (not limited by page_size).

Page number echoed from the request. Aggregation buckets are not paginated - the full set of buckets for the date range is always returned regardless of page.

Total number of pages implied by total_hits and page_size, for reference only - the aggregations list itself is not split across pages.

Page size echoed from the request. Used only to compute total_pages; does not limit the number of aggregation buckets returned.