This project is for my paper: [Mining User Similarity from GPS Trajectory Based on Spatial-temporal and Semantic Information] (https://ieeexplore.ieee.org/document/9874192/metrics#metrics)
ST_similarity This file includes location features (could include point features ellipse features in the future) Location features compute the tfidf vector of the clusters of stay points.
Semc_similarity Semc firstly generate the semantic sequences from stay region clusters, then use prefixspan compute the maximal frequent sequence patterns, finally compute the similarity of mfsp from every 2 users as their semantic similarity.
More details please refer to the paper. If you wanna use this code, please refer this paper in your research properlly.