Download now Free registration required
Mobile devices can produce continuous streams of data which are often specific to the person carrying them. The authors show that cell phone tracks from the MIT Reality dataset can be used to reliably characterize individual people. This is done by treating each person's data as a separate language by building a standard n-gram language model for each "Author." They, then compute the perplexities of an unlabelled sample as based on each person's language model. The sample is assigned to the user yielding the lowest perplexity score. This technique achieves 85% accuracy and can also be used for clustering. They also show how language models can also be used for predicting movement and propose metrics to measure the accuracy of the predictions.
- Format: PDF
- Size: 147.5 KB