Metasearch infrastructure with incremental updates

摘要

Techniques are provided for a metasearch infrastructure. The metasearch infrastructure supports one or more of the following: hotel matching, backend support of user queries, frontend support for user queries, a cache layer that groups or organizes cached data in storage based on hotel location, scoring and ranking on the client side, experiments and rankers, duplicate booking prevention, blacklisting hotels, and indexing that utilizes a document-independent data structure to store index values.

主权项

1. A computer-implemented method comprising:
receiving, from a client, at a server system operated by a first party, a first client request for hotel data that satisfies one or more criteria, wherein the first client request includes a first plurality of hotel identifiers, each of which uniquely identifies a different hotel from a plurality of hotels; identifying a mapping that maps a first set of identifiers to a second set of identifiers that were established by one or more other parties, each of which is different than the first party and stores specific hotel information in association with an identifier in the second set of identifiers; using the mapping to identify a second plurality of hotel identifiers based on the first plurality of hotel identifiers, wherein each hotel identifier in the second plurality of hotel identifiers (a) uniquely identifies a different hotel from the plurality of hotels; (b) was established by a second party that is different than the first party; and (c) is different than the corresponding hotel identifier in the first plurality of hotel identifiers; in response to receiving the first client request, generating a plurality of requests, each of which includes a different subset of the second plurality of hotel identifiers and requests a different set of data from a first source that is (1) remote relative to the server system and (2) operated by the second party that is different than the first party that operates the server system; sending the plurality of requests over a network from the server system to the first source of the second party; after sending the plurality of requests to the first source, receiving, at the server system, a plurality of responses from the first source; sending, from the server system, to the client, data from the plurality of responses; wherein the method is performed by one or more computing devices.