Data indexing system, data indexing method and data querying method

摘要

A data indexing system is provided. The system includes: a time interval splitting unit configured to split a predetermined time range into a plurality of time intervals. A first index creating unit is configured to create a first-level index for data in a current time interval, the first-level index including a time interval index and a subspace index. A second index creating unit is configured to create, in a time interval after the current time interval, a second-level index for data in the current time interval. The second-level index has a finer granularity than the first-level index. A data indexing method and a data querying method are also provided. The cost for index maintenance can be greatly reduced in a case of frequent updates. Further, invalid scans during the query process can be reduced and the query performance can be improved.

主权项

1. A data indexing system, executed on a computing device, comprising:
a time interval splitting unit configured in computer readable storage media to split a predetermined time range into a plurality of time intervals, the time interval splitting unit is configured: to split a predetermined tim me intervals based upon a priori knowledge such that the data is distributed uniformly in the plurality of time intervals, to calculate a uniformity of data splitting and a number of data layers for each time interval; to split a certain time interval into more time intervals if the uniformity of data splitting for the certain time interval is larger than a first threshold and the number of data layers for the certain time interval is larger than a second threshold, or to combine a certain time interval with an adjacent time interval if the uniformity of data splitting for the certain time interval is less than the first threshold and the number of data layers for the certain time interval is less than the second threshold; a first index creating unit configured to create a first-level index for data in a current time interval, the first-level index including a time interval index and a subspace index; and a second index creating unit configured to create, in a time interval after the current time interval, a second-level index for data in the current time interval, wherein the second-level index has a finer granularity than the first-level index.