Create README.md

Author: PanSours

README.md

@@ -0,0 +1,41 @@
+# Synopsis
+This project enables to study the problem of finding similarity between trajectories (trajectory similarity problem) which is a known problem in the field of Data Mining. The algorithm called to deploy and solves the above problem is the LCSS(Longest Common Subsequence). Each orbit is described by a sequence geographical points (latidute, longitude). Each record dataset that we are given represents a path of a taxi that has taken place in the city of Beijing. Each line of the file contains the following separated by commas: taxi id, timestamp, latitude, longitude.
+
+# Background
+Each record dataset that we are given represents a path of a taxi that has taken place in the city of Beijing. One illustrative example of the format of the file is:
+
+366, Mon Mar 03 00:05:59 EET 2014,39.90732,116.45353
+366, Mon Mar 03 00:10:59 EET 2014,39.90729,116.45348
+366, Mon Mar 03 00:15:59 EET 2014,39.90725,116.45334
+366, Mon Mar 03 00:20:59 EET 2014,39.90722,116.4533
+366, Mon Mar 03 00:25:59 EET 2014,39.90722,116.45327
+366, Mon Mar 03 00:30:59 EET 2014,39.90725,116.4532
+366, Mon Mar 03 00:35:59 EET 2014,39.9076,116.45309
+366, Mon Mar 03 00:40:59 EET 2014,39.9077,116.453
+366, Mon Mar 03 00:45:59 EET 2014,39.9076,116.45281
+366, Mon Mar 03 00:50:59 EET 2014,39.90767,116.45271
+366, Mon Mar 03 00:55:59 EET 2014,39.90771,116.45262
+
+Each line of the file contains the following separated by commas:
+taxi id, timestamp, latitude, longitude. The features we are usings are  the latitude (geographical
+width of a point) and the longitude (longitude of a point).
+
+In the project we have implemented the following:
+
+1. Implementation of the algorithm LCSS and support of the implementation by using a window environment
+corresponding framework (JavaFx).
+
+2. We have modified our program so that having as input two trajectories S, Q with lengths
+Ls, Lq and Ls >> Lq, to find the subset of S trajectory, which presents the greatest resemblance to the Q trajectory comparison. That is to
+returning the S portion of length: Lq +  d (0 <d <Lq) having the largest percent similarity with Q.
+
+3. Map Display (Google Maps) of the track comparison, in conjunction with
+the most common orbit found through the LCSS.
+
+# Installation
+Platform : NetBeans IDE 7.4 or greater, Java version : 7 or greater.
+All you have to do is open the project in Netbeans IDE as a JavaFx project.
+
+# Running the project
+You can run the application by running the MainWindow.java file. Then you have to select for which datasets you want to run the Lcss algorithm and hit the Procedure button.
+If you want to see the trajectories on the map, you can hit the Google Maps button.