Continuous Trajectory Similarity Search for Online Outlier Detection

The code officially implements the CTSS algorithm for online outlier detection.

Supported query modes

1. Single point query in Euclidean space
2. Continuous query in Euclidean space
3. Single point query in Road network space
4. Continuous query in Road network space

File format

Vertex file

when reading the vertex in the file, the structure of each line is required to be:

VertexID	Latitude	Longitude

Example:

0	39.9476009	116.4020469
1	39.9476591	116.4053914
2	39.9319029	116.4025943
...

Edge file

when reading the edge in the file, the structure of each line is required to be:

EdgeID	FirstVertexID	SecondVertexID

Example:

0	0	1
1	2	701
2	701	3
...

Reference path file

when reading the reference path in the file, the structure of each line is required to be:

EdgeID	FirstVertexID	SecondVertexID

Example:

1578	1902	1201
2764	1201	114
2765	114	302
...

Complete path file

when reading the complete path in the file, the structure of each line is required to be:

EdgeID	FirstVertexID	SecondVertexID

Example:

1578	1902	1201
2764	1201	114
2765	114	302
...

Build

make

Run

./CTSS "input_folder_path" "output_folder_path" "start_position" "stop_position" "delta" "query_mode" "section_length"

Example:

./CTSS data/beijing/set_14000 data/beijing/out 1 20 0.0081 1 100

Contributors

• Zhihao Chang: changzhihao@zju.edu.cn

Reference

If you want to use this code, please cite:

@article{DBLP:journals/tkde/ZhangCWYTC22,
  author       = {Dongxiang Zhang and
                  Zhihao Chang and
                  Sai Wu and
                  Ye Yuan and
                  Kian{-}Lee Tan and
                  Gang Chen},
  title        = {Continuous Trajectory Similarity Search for Online Outlier Detection},
  journal      = {{IEEE} Trans. Knowl. Data Eng.},
  volume       = {34},
  number       = {10},
  pages        = {4690--4704},
  year         = {2022},
}