Tianbaojie / reproduce_LUAD

Parameter	Definition	Default	Remarks
	input_path	Path of trace data	./input
	output_path	Path of output	./output
	save_weights	Save candidates' weight or not	False
	region_centre	Location of regions' centers	./RegionCenters
	region_popularity	Popularity of regions	./RegionPopularity
	Tmax	The number of time-bins	24*8
	scale	Spatial granularity	0.01
	K	Top-K in hit-precision	10

1. Run

   $ python algorithms/pois.py --input_path ./algorithms/input/ --output_path ./algorithms/output/ --save_weights False --region_centre ./algorithms/RegionCenters --region_popularity ./algorithms/RegionPopularity --scale 0.01 --K 10

   $ python algorithms/pois.py \
    --input_path ./algorithms/input/ \
    --output_path ./algorithms/output/ \
    --save_weights False \
    --region_centre ./algorithms/RegionCenters \
    --region_popularity ./algorithms/RegionPopularity \
    --scale 0.01 \
    --K 10

数据格式

数据分为fs_example.txt和tt_example.txt分别是foursquare和twitter的轨迹数据，每一行代表一个用户的轨迹。每一行都是如下格式

101;|3836,963221636|4053,577522790|

分号前的字符串是用户 ID。轨迹中的时空点用竖线划分。逗号前的元素是时间的ID，逗号后的元素是binId。

REFERENCES

[WWW 2016] C. Riederer, Y. Kim, A. Chaintreau, N. Korula, and S. Lattanzi, “Linking users across domains with location data: Theory and validation,” in Proc. WWW, 2016.