Literature on human mobility modeling and prediction

In this repository you will find the metadata of all the relevant literature concerning human-mobility modeling and prediction. The repository also contains the source codes for scrapping, cleaning and preprocessing this data.

From where is this data collected?

• Web of Science
• Microsoft Academic Search
• Semantic Scholar

How is the scraping and crawling done?

• The scrapping from Semantic Scholar is done using Semantic Scholar API.
• The maka Python module is used to scrape data from Microsoft Academic Search.
• Web of Science permits to manually download data for selected topics under a university license.

What key-words and domains were selected to scrape this data?

The scrapping was done using the following Keywords:

• Human-mobility prediction
• Human-mobility modeling
• Next place forecasting
• Next place prediction
• Predicting User Movement
• Predicting Significant locations
• Forecasting Next User Location

and restricted to the following domains:

• [Computer science, Computer network, Mobility model, Distributed computing, Real-time computing, Wireless ad hoc network]
• [Handover, Artificial intelligence, Optimized Link State Routing Protocol, Machine learning, Quality of service]
• [Wireless Routing Protocol, Data mining, KDD, Wireless network, Mobile ad hoc network, Ad hoc wireless distribution service]
• [Mobile computing, Mobility management, Adaptive quality of service multi-hop routing]

How does the pipeline for the collection and cleaning process look like?

How was the cleaning, duplicate removal and unrelated item removal was done?

• The string similarity and profiling was performed using python-string-similarity.
• The raw_data folder contains raw data scrapped from all the platforms with the title indicating the key word searched.
• The scripts folder contains python scripts used to scrape, preprocess, clean and merge the data collected from different sources.
• The file literature_metadata contains the merged list and literature_with_abstracts contains all the abstracts.
• The abstract parsing to identify the techniques is based on Automated Keyword Extraction from Articles

What is the format of the data?

Paper Title | Publication Year | Publication Venue | Authors | Keyword List

Dataset Structure: GeoLife

Dataset Structure: PrivaMov

Dataset Structure: MDC