Information processing device

• US 9,292,555 B2
• Filed: 03/15/2012
• Issued: 03/22/2016
• Est. Priority Date: 04/05/2011
• Status: Active Grant

• Current Assignee: Cloud BYTE LLC
• Original Assignee: NEC Corporation
• Inventors: Ogura, Shoji
• Primary Examiner(s): Ly, Anh
• Application Number:
  US14/009,778
• Publication Number:
  US 20140129565A1
• Time in Patent Office: 1,468 Days
• Field of Search: 707/741, 707/761, 370/222, 370/392, 370/401, 370/404, 370/463, 370/245, 370/254, 370/223, 370/224, 370/232, 370/236, 370/231, 370/340, 345/419, 345/426, 345/424, 345/440, 345/441, 715/823, 715/802, 715/715, 715/852, 715/798, 715/801, 715/800, 715/810, 716/4, 716/18, 716/7, 716/6, 716/14, 716/15, 716/16, 703/4, 703/5, 703/13, 703/14, 703/15, 73/32
• US Class Current: 1/1
• CPC Class Codes: G06F 16/2228 Indexing structures
• CPC Class Codes: G06F 16/41 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/901 Indexing; Data structures therefor; Storage structures for retrieval from the web G06F16/951
• CPC Class Codes: H04L 41/0826 for reduction of network costs H04L41/0833 takes precedence
• CPC Class Codes: H04L 67/1004 Server selection for load balancing
• CPC Class Codes: H04L 67/1012 based on compliance of requirements or conditions with available server resources
• CPC Class Codes: H04L 67/1023 based on a hash applied to IP addresses or costs

Abstract

An information processing device of the present invention includes: an index generating means for generating index data distributedly stored into a plurality of servers, each index data being generated for each data point information including metadata representing a characteristic of data to be searched, and each index data being associated with reference point information that are other data point information such that distances to the each data point information based on a preset reference are within a preset range; and a communication cost storing means for storing, for each of the servers storing the index data, communication cost information representing a preset cost that costs in communication with the server. The index generating means selects the reference point information to be associated with specific one of the data point information, based on the communication cost information of the server storing the index data corresponding to the reference point information.

Claims

1. An information processing device comprising:

an index generating unit for generating index data that are distributedly stored into a plurality of servers, respectively, each of the index data being generated for each of data point information including metadata representing a characteristic of data to be searched, and each of the index data being associated with reference point information that are others of the data point information such that distances to the each data point information based on a previously set reference are within a previously set range; and

a communication cost storing unit for storing, for each of the servers storing the respective index data, communication cost information representing a previously set cost that costs in communication with the server,

wherein the index generating unit is configured to select the reference point information to be associated with specific one of the data point information, based on the communication cost information of the server storing the index data corresponding to the reference point information, and

wherein the index generating unit is configured to;

exclude one of reference point candidate information from among the reference point candidate information that are the others of the data point information to become candidates of the reference point information for the specific one of the data point information and calculate, with reference to one of remainder of the reference point candidate information, a coordinate point such that coordinates are distances of others of the reference point candidate information;

calculate a volume of a polyhedron such that each of vertexes is the coordinate point calculated with reference to each of the remainder of the reference point candidate information; and

select the reference point information to be associated with the specific one of the data point information, based on the calculated volume and the communication cost information of the server storing the index data corresponding to the excluded one of the reference point candidate information.

2. The information processing device according to claim 1, wherein the index generating unit is configured to select a previously set number of the reference point information to be associated with the specific one of the data point information, based on the communication cost information of the server storing the index data corresponding to the reference point information.

3. The information processing device according to claim 1, wherein the index generating unit is configured to select the reference point information by excluding, from among the reference point information to be associated with the specific one of the data point information, the excluded one of the reference point candidate information such that a value obtained by adding the communication cost information of the server storing the index data corresponding to the excluded one of the reference point candidate information to the calculated volume is larger than any values obtained when others of the reference point candidate information are excluded, respectively.

4. The information processing device according claim 1, wherein:

the communication cost storing unit is configured to store communication cost information representing a previously set cost that costs in communication between the servers, as the communication cost information for each of the servers storing the index data; and

the index generating unit is configured to select the reference point information to be associated with the specific one of the data point information, based on the communication cost information between the server storing the index data corresponding to the specific one of the data point information and the server storing the index data corresponding to the other of the data point information to become the candidates of the reference point information.

5. The information processing device according to claim 4, wherein the communication cost storing unit is configured to store information representing a previously set cost that costs in communication between the servers and that occurs depending on a difference in processing capability between the abovementioned servers, as the communication cost information for each of the servers storing the index data.

6. An information processing device comprising:

an index generating means unit for generating index data that are distributedly stored into a plurality of servers, respectively, each of the index data being generated for each of data point information including metadata representing a characteristic of data to be searched, and each of the index data being associated with reference point information that are others of the data point information such that distances to the each data point information based on a previously set reference are within a previously set range;

a communication cost storing unit for storing, for each of the servers storing the respective index data, communication cost information representing a previously set cost that costs in communication with the server;

an index storing unit for storing the index data corresponding to predetermined one of the data point information;

a search request accepting unit for accepting a request for search of the data to be searched;

a searching unit for searching the data point information corresponding to the search request accepted by the search request accepting unit in accordance with a previously set reference, from the index data stored in the index storing unit; and

a search request transferring unit for transferring the search request to another information processing device in a case where the searching unit determines that the data point information corresponding to the search request accepted by the search request accepting unit in accordance with the previously set reference does not exist in the index data stored in the index storing means unit,

wherein the index generating unit is configured to select the reference point information to be associated with specific one of the data point information, based on the communication cost information of the server storing the index data corresponding to the reference point information.

7. A non-transitory computer-readable medium storing a computer program comprising instructions for:

causing an information processing device to realize an index generating unit, wherein the information processing device includes a communication cost storing unit for storing, for each of servers storing respective index data, communication cost information representing a previously set cost that costs in communication with the server, and the index generating unit is a unit for generating index data that are distributedly stored into a plurality of servers, respectively, each of the index data being generated for each of data point information including metadata representing a characteristic of data to be searched, and each of the index data being associated with reference point information that are others of the data point information such that distances to the each of the data point information based on a previously set reference are within a previously set range; and

causing the index generating unit to select the reference point information to be associated with specific one of the data point information, based on the communication cost information of the server storing the index data corresponding to the reference point information,

wherein the index generating unit is configured to;

exclude one of reference point candidate information from among the reference point candidate information that are the others of the data point information to become candidates of the reference point information for the specific one of the data point information and calculate, with reference to one of remainder of the reference point candidate information, a coordinate point such that coordinates are distances of others of the reference point candidate information;

calculate a volume of a polyhedron such that each of vertexes is the coordinate point calculated with reference to each of the remainder of the reference point candidate information; and

select the reference point information to be associated with the specific one of the data point information, based on the calculated volume and the communication cost information of the server storing the index data corresponding to the excluded one of the reference point candidate information.

8. An information processing method in an information processing device including a communication cost storing unit for storing, for each of servers storing index data, communication cost information representing a previously set cost that costs in communication with the server, the information processing method comprising:

generating index data that are distributedly stored into a plurality of servers, respectively, each of the index data being generated for each of data point information including metadata representing a characteristic of data to be searched, and each of the index data being associated with reference point information that are others of the data point information such that distances to the each of the data point information based on a previously set reference are within a previously set range, selecting the reference point information to be associated with specific one of the data point information, based on the communication cost information of the server storing the index data corresponding to the reference point information,

wherein when generating the index data the method further comprises;

excluding one of reference point candidate information from among the reference point candidate information that are the others of the data point information to become candidates of the reference point information for the specific one of the data point information and calculating, with reference to one of remainder of the reference point candidate information, a coordinate point such that coordinates are distances of others of the reference point candidate information;

calculating a volume of a polyhedron such that each of vertexes is the coordinate point calculated with reference to each of the remainder of the reference point candidate information; and

selecting the reference point information to be associated with the specific one of the data point information, based on the calculated volume and the communication cost information of the server storing the index data corresponding to the excluded one of the reference point candidate information.