safarzadeh-reza / KDD2021_guizu

Source codes of KDD2021 paper "Generating Mobility Trajectories with Retained Data Utility"

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Source codes of SIGKDD 2021 paper:

"Generating Mobility Trajectories with Retained Data Utility"

There are FIVE folders in total:

  • Data Samples

This folder contains the Mobility Data Samples used in this paper. The example data is in the folder named "Original Data Samples". At the same time, we also provide mobility data samples generated by TrajGen, which can be found in the folder named "TrajGen Generated Samples".

  • DCGAN Model

This folder includes the implementation of DCGAN paper, which is also the model used in TrajGen.

  • Location - Image Translation

This folder gives the codes of translation between locations and images. The codes can be used to convert locations in trajectories into images. You can also use them to conduct the translation from images to locations.

  • Map Matching

An implementation of HMM (Hidden Markov Model) based Map Matching algorithm is provided in this folder. The details can be found in the original paper.

  • Seq2Seq Model

This folder has the code of Seq2Seq model, which is used to recover the correct sequence information of our newly generated locations. It is an implementation of attention mechanism involved encoder-decoder model. You might find the details of the model in the original paper.

About

Source codes of KDD2021 paper "Generating Mobility Trajectories with Retained Data Utility"


Languages

Language:Python 100.0%