Learning space-time index method, device and medium based on global interval error
A learning-based, space-time technology, applied in the field of big data, can solve problems such as insufficient query efficiency and complex index structure, and achieve high-precision, fast retrieval, and improved accuracy
Active Publication Date: 2022-02-18
ZHEJIANG UNIV
View PDF12 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0006] The purpose of the present invention is to overcome the problems of complex index structure and insufficient query efficiency in the existing index technology in the context of massive spatio-temporal data, and to provide a learning-type spatio-temporal index method, device and medium based on global interval error
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0105] This example uses real New York taxi data as experimental data to explore the advancement of the method in real data scenarios. This data set is called D 9 . Data set D 9 The total amount of data is 5,000,000, the data space range is [-74.2605,40.4964,-73.7347,40.9192], and the time range is [1396281600000,1398873597000].
[0106] According to the aforementioned steps S1 and S2, aggregation and quantile conversion (ie, QM conversion) are performed, and the original data set D 9 Longitude x ,latitude y and time t The coordinate values under the three coordinate dimensions are aggregated and arranged in order to form three monotone ordered sequences, and then each monotone ordered sequence is mapped to the uniformly distributed data space by QM transformation, and the three coordinate dimensions are obtained. Evenly distributed sequence. Among them, the number of quantiles N p and monotonic ordered sequence O' k The same length, that is, a monotonic ordered se...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a learning type space-time index method, device and medium based on the global interval error. The index method is divided into two processes: index model construction and index model-based range retrieval, and is a multi-directional learning spatio-temporal index. When building an index model, it is necessary to aggregate repeated data, and then convert the data into a uniform distribution, and then train the index model; when performing range retrieval based on the index model, it is necessary to convert the retrieval range to a uniformly distributed data space, and then independently Search the location distribution range, and finally compare the location distribution ranges of each dimension to determine the final multi-dimensional retrieval results. Compared with the common space-time tree index, this index method has better space-time range retrieval performance, especially the repeated data aggregation operation and data space conversion operation in the index model construction process, which have effectively improved its superiority.
Description
technical field [0001] The invention relates to the technical field of big data, in particular to a spatio-temporal index and data retrieval method in the high-performance storage field of geographic spatio-temporal big data. Background technique [0002] Common single-node spatiotemporal indexes are divided into two types: grid indexes based on spatial division and data-driven tree indexes. [0003] The grid-like index divides the overall spatial area according to predetermined rules to form a grid system. Each grid unit is assigned a unique number, and the data index is realized by one-to-one correspondence between spatio-temporal objects and grid coordinates. It is the earliest type of spatial index. The implementation of grid index is simple and the query efficiency is high, but it will cause great data redundancy and cannot be directly applied to the scene of large-scale spatio-temporal data, and when the spatio-temporal distribution of data is unbalanced, each grid uni...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/901G06F16/9537G06N3/04G06N3/08
CPCG06F16/9027G06F16/9537G06N3/04G06N3/08
Inventor 胡林舒张丰陈宁华覃梦娇汪愿愿吴森森杜震洪傅晨华
Owner ZHEJIANG UNIV