An aggregation service aggregates discussions of events and topics within a given subject area using seed search queries, seed URLs, and/or search engine data. A subject area is selected and a set of seed search queries and/or seed URLs relevant to that subject area are manually generated. The set of seed search queries and/or seed URLs are used to identify URLs containing content relevant to the subject area. The URLs are crawled at given intervals to identify content items, which are analyzed using a classifier to identify content items relevant to the subject area. The content items are grouped into clusters surrounding events and/or topics within the subject area. The various clusters are ranked against one another to facilitate ordering clusters for presentation to end-users.