Method for selecting a subset of content sources from a collection of content sources
A method for selecting a subset of content sources from a collection of content sources is disclosed. A server retrieves, in response to a plurality of queries on a topic from a client, using a programmed computer, a plurality of sets of documents from the collection of content sources. The server enumerates all subsets of the plurality of sets of documents. The server calculates, for each subset, a count of effectiveness of a subset and a price of the subset. The server selects a subset having the highest calculated ratio of count of effectiveness of the subset to price of the subset. The server delivers the selected subset of the plurality of sets of documents to the client.
1. A method, comprising:
retrieving, in response to a plurality of queries on a topic from a client, by a programmed computer, a plurality of sets of documents from a collection of content sources;
enumerating, by the programmed computer, all subsets of documents of the plurality of sets of documents corresponding to all subsets of the collection of content sources;
calculating, by the programmed computer, for each subset of the collection of content sources, a count of effectiveness of a corresponding subset of documents and a price of the corresponding subset of documents, wherein said calculating the count of effectiveness of the corresponding subset of document comprises calculating, for each enumerated subset of documents, an equivalent fractional story count, wherein an exact duplicate between two documents of an enumerated subset of documents is counted as a normalized equivalent fractional story count of one, and wherein one document of an exact duplicate between two documents of the enumerated subset of documents is discarded from the subset of documents;
selecting a subset of content sources having the highest calculated ratio of count of effectiveness of the corresponding subset of documents to price of the corresponding subset of documents; and
delivering the selected subset of content sources and the corresponding subset plurality of sets of documents to the client.
2. The method of claim 1 , wherein the retrieved plurality of sets of documents from the collection of content sources includes only documents having relevance above a predetermined threshold.
3. The method of claim 1 , wherein the equivalent fractional story count is based on a degree of duplication between documents of an enumerated subset of documents.
4. The method of claim 1 , wherein the equivalent fractional story count is based on normalized word count overlaps between two documents of an enumerated subset of documents.
5. The method of claim 4 , wherein normalized word count overlaps are based on Jaccard Counts.
6. The method of claim 1 , wherein a partial duplication between two documents of an enumerated subset of documents is counted as a normalized equivalent fractional story count between zero and one.
7. The method of claim 1 , wherein the count of effectiveness of a selected subset of content sources is a difference between the total number of documents in the corresponding subset of documents and a sum of equivalent fractional story counts of the corresponding subset of documents.
8. The method of claim 1 , wherein the calculated price of a corresponding subset of documents is a sum of individual prices of documents in the corresponding subset of documents.
9. The method of claim 1 , wherein at least one subset of documents of the sets of documents changes over time and is updated.
10. The method of claim 9 , wherein the at least one subset of content sources includes newsfeeds, wherein the historical end of the newsfeeds continually moves forward in time.
11. The method of claim 9 , wherein the at least one subset of content sources includes evergreen, wherein new content is added but old content is never purged.
12. The method of claim 9 , wherein the at least one subset of content sources of the sets of documents is archival and is never updated.
13. The method of claim 1 , wherein the plurality of queries is assembled from a list of names or authorities used to probe each of the sets of documents.
14. The method of claim 13 , wherein the plurality of queries used to probe each of the sets of documents is continually changed over time.
15. The method of claim 13 , wherein the plurality of queries is taken as a set, but run individually, and the answers pooled to form one set of results for analysis.
16. The method of claim 15 , wherein duplication elimination, exact or approximate, and subsequent document count reduction, in whole units or fractions, are performed for each query in the set individually before merging the answers across all queries.
17. The method of claim 1 , wherein the plurality of queries is assembled from records of actual queries submitted by users, as recorded in query log files.
18. The method of claim 17 , the plurality of queries is one of a sum of all queries over a predetermined time period, a subset of queries chosen at random, or a subset of queries chosen by importance-bias sampling.
19. The method of claim 1 , wherein a subscription set is adjusted automatically as the selected subset of the plurality of sets of content sources changes.
20. The method of claim 1 , wherein the set of content sources, the types of queries in the plurality of queries, and a content budget are managed in combination as any one changes over time.
21. A non-transitory computer readable storage medium including instructions that, when executed by a processing system, cause the processing system to perform a method operations comprising:
retrieving, in response to a plurality of queries on a topic from a client, by a programmed computer, a plurality of sets of documents from a collection of content sources;
enumerating, by the programmed computer, all subsets of documents of the plurality of sets of documents corresponding to all subsets of the collection of content sources;
calculating, by the programmed computer, for each subset of the collection of content sources, a count of effectiveness of a corresponding subset of documents and a price of the corresponding subset of documents, wherein said calculating the count of effectiveness of the corresponding subset of document comprises calculating, for each enumerated subset of documents, an equivalent fractional story count, wherein an exact duplicate between two documents of an enumerated subset of documents is counted as a normalized equivalent fractional story count of one, and wherein one document of an exact duplicate between two documents of the enumerated subset of documents is discarded from the subset of documents;
selecting a subset of content sources having the highest calculated ratio of count of effectiveness of the corresponding subset of documents to price of the corresponding subset of documents; and
delivering the selected subset of content sources and the corresponding subset plurality of sets of documents to the client.
22. A system, comprising:
a memory;
a processing device, coupled to the memory, the processing device to:
retrieve, in response to a plurality of queries on a topic from a client a plurality of sets of documents from a collection of content sources;
enumerate all subsets of documents of the plurality of sets of documents corresponding to all subsets of the collection of content sources;
calculate, for each subset of the collection of content sources, a count of effectiveness of a corresponding subset of documents and a price of the corresponding subset of documents, wherein said calculate the count of effectiveness of the corresponding subset of document comprises calculate, for each enumerated subset of documents, an equivalent fractional story count, wherein an exact duplicate between two documents of an enumerated subset of documents is counted as a normalized equivalent fractional story count of one, and wherein one document of an exact duplicate between two documents of the enumerated subset of documents is discarded from the subset of documents;
select a subset of content sources having the highest calculated ratio of count of effectiveness of the corresponding subset of documents to price of the corresponding subset of documents; and
deliver the selected subset of content sources and the corresponding subset of documents to the client.