Model BuilderΒΆ

The model builder is a preprocessing stage that identifies sources of technical inconsistencies and provides stochastic information about distribution of various configurations. As the name implies, this tool builds the model file based on the pseudocode provided in Listing 1 in the paper. To build the model, Model Builder takes as input a dataset of web fingerprints as well as a Uniqueness Threshold, builds the model and saves it as a cPickle file. This file is later used in the other modes make an inference based on the input.

Example usage and output:

$ ./main.py -m fingerprint 100 ./model_files/extracted_data_freq100.pickle

Username: root
Password:
List of attributes in the dataset:
counter, id, addressHttp, time, userAgentHttp, acceptHttp, hostHttp, connectionHttp, encodingHttp, languageHttp, orderHttp, pluginsJS, platformJS, cookiesJS, dntJS, timezoneJS, resolutionJS, localJS, sessionJS, IEDataJS, canvasJS, webGLJs, fontsFlash, resolutionFlash, languageFlash, platformFlash, adBlock, vendorWebGLJS, rendererWebGLJS, octaneScore, sunspiderTime, pluginsJSHashed, canvasJSHashed, webGLJsHashed, fontsFlashHashed
Please input the irrelevant parameters in the dataset separated by a comma or press Enter to continue:octaneScore, sunspiderTime, timezoneJS, id, counter, time
5.000000% of columns are done...
11.000000% of columns are done...
14.000000% of columns are done...
17.000000% of columns are done...
20.000000% of columns are done...
22.000000% of columns are done...
25.000000% of columns are done...
28.000000% of columns are done...
31.000000% of columns are done...
34.000000% of columns are done...
37.000000% of columns are done...
40.000000% of columns are done...
45.000000% of columns are done...
48.000000% of columns are done...
51.000000% of columns are done...
54.000000% of columns are done...
57.000000% of columns are done...
60.000000% of columns are done...
62.000000% of columns are done...
65.000000% of columns are done...
68.000000% of columns are done...
71.000000% of columns are done...
74.000000% of columns are done...
77.000000% of columns are done...
80.000000% of columns are done...
88.000000% of columns are done...
91.000000% of columns are done...
94.000000% of columns are done...
97.000000% of columns are done...
Model with uniqueness threshold of 100 was built in 50 seconds.