Methods for effective processing of time series

• US 9,047,349 B2
• Filed: 04/02/2010
• Issued: 06/02/2015
• Est. Priority Date: 04/05/2009
• Status: Active Grant

• Current Assignee: Tamiras Per Pte Ltd LLC (Intellectual Ventures LLC)
• Original Assignee: Yanicklo Technology LLC (Intellectual Ventures LLC)
• Inventors: Shaked, Guy, Braverman, Vladimir, Rubin, Gabby, Belyaev, Victor, Sadetsky, Marina
• Primary Examiner(s): Hershley, Mark E
• Application Number:
  US12/753,388
• Publication Number:
  US 20110060753A1
• Time in Patent Office: 1,887 Days
• Field of Search: 707/752
• US Class Current: 1/1
• CPC Class Codes: G06F 16/2477 Temporal data queries
• CPC Class Codes: G06F 16/284 Relational databases
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

A method of effectively representing and processing data sets with time series is disclosed. The method may comprise representing time series as a virtual part of data in a data store layer of a user system, thereby allowing processing of time-series related queries in said data store layer of said user system.

Claims

1. A method comprising:

maintaining a data store that is configured to provide an answer to one or more queries comprising a requested dimension and a requested measure;

representing at least a portion of time series data as a virtual part of data in the data store without actually being stored in the data store, the time series data comprising a sequence of data elements given in some order of a time attribute, thereby enabling processing of time-series related queries in a data store layer of said user system;

organizing data having a plurality of dimensions, including a time dimension and a non-time dimension, in the data store according to at least a first of said one or more dimensions, said first dimension being other than the time dimension, wherein the data is not stored in timestamp order, to form a first plurality of data streams for respective objects in the data, followed by organizing data in respective data streams for respective objects according to the time dimension, to thereby split the data into disjoint chunks, with respect to the time dimension, which can be processed separately in response to a time series query;

receiving, at a computer system comprising one or more computing devices, at least a first time series query from a user directed to the data store; and

at least partly in response to receiving the first time series query and without translating the first time series query into middle queries, scanning the data store layer, and manipulating extracted data entries to generate, by the computer system, a response to the first time series query at the data store layer based at least in part on data from the time series data that is a virtual part of data in the data store that is not actually stored in the data store.

2. The method of claim 1, where said computer system further comprises a relational database.

3. The method of claim 1, wherein the time series data relates to stocks.

4. A computing system comprising:

a processor; and

a computer readable medium storing machine-executable instructions including one or more modules configured for execution by the processor in order to cause the computing system to;

maintain a data store that is configured to provide an answer to one or more queries comprising a requested dimension and a requested measure;

represent at least a portion of time series data as a virtual part of data in the data store without actually being stored in the data store, the time series data comprising a sequence of data elements given in some order of a time attribute, thereby enabling processing of time-series related queries in a data store layer of said user system;

organize data having a plurality of dimensions, including a time dimension and a non-time dimension, in the data store according to at least a first of said one or more dimensions, said first dimension being other than the time dimension, wherein the data is not stored in timestamp order, to form a first plurality of data streams for respective objects in the data, followed by

organize data in respective data streams for respective objects according to the time dimension, to thereby split the data into disjoint chunks, with respect to the time dimension, which can be processed separately in response to a time series query;

receive at least a first time series query from a user directed to the data store; and

at least partly in response to receipt of the first time series query, and without translating the first time series query into middle queries, scan the data store layer, and manipulate extracted data entries to;

generate a response to the first time series query at the data store layer based at least in part on data from the time series data that is a virtual part of data in the data store that is not actually stored in the data store.

5. The computing system of claim 4, where said computing system further comprises a relational database.

6. The computing system of claim 4, wherein said computing system is a complementary system, which is added onto an existing user system using an existing interface.

7. The computing system of claim 4, wherein said computing system is an integrated system, which is integrated into an existing user system.

8. The computing system of claim 7, wherein said computing system is a complementary system, which is added onto an existing user system using an existing interface.

9. The computing system of claim 4, wherein said computing system is an integrated system, which is integrated into an existing user system.

10. The computing system of claim 4, where computing user system comprises a relational database.

11. The computing system of claim 4, wherein the time series data relates to stocks.

12. The computing system of claim 4, wherein the time series data relates to shipped orders from an on-line store.

13. The computing system of claim 4, wherein the time series data becomes part of a data entry for the data store.

14. A non-transitory computer-readable medium encoded with instructions thereon, wherein the instructions are readable by a computing system in order to cause the computing system to perform operations comprising:

maintaining a data store that is configured to provide an answer to one or more queries comprising a requested dimension and a requested measure;

representing at least a portion of time series data as a virtual part of data in the data store without actually being stored in the data store, the time series data comprising a sequence of data elements given in some order of a time attribute, thereby enabling processing of time-series related queries in a data store layer of said user system;

organizing data having a plurality of dimensions, including a time dimension and a non-time dimension, in the data store according to at least a first of said one or more dimensions, said first dimension being other than the time dimension, wherein the data is not stored in timestamp order, to form a first plurality of data streams for respective objects in the data, followed by organizing data in respective data streams for respective objects according to the time dimension, to thereby split the data into disjoint chunks, with respect to the time dimension, which can be processed separately in response to a time series query;

receiving at least a first time series query from a user directed to the data store; and

at least partly in response to receiving the first time series query, and without translating the first time series query into middle queries, scanning the data store layer, and manipulating extracted data entries to generate a response to the first time series query at the data store layer based at least in part on data from the time series data that is a virtual part of data in the data store that is not actually stored in the data store.

15. The non-transitory computer-readable medium of claim 14, wherein the time series data relates to stocks.