Changeset 68 for trunk/EventBenchConsole/src/de/ugoe/cs/eventbench/web/WeblogParser.java

Timestamp:

06/15/11 16:54:11 (13 years ago)

Author:

sherbold

Message:

• changed weblog parser such that session without a minimum length are pruned to remove single page visits from the sessions (e.g., through robots)
• changed loadSessionsFromClickstream such that the session timeout and the minimum session length are optional parameters

File:

• trunk/EventBenchConsole/src/de/ugoe/cs/eventbench/web/WeblogParser.java (modified) (3 diffs)

trunk/EventBenchConsole/src/de/ugoe/cs/eventbench/web/WeblogParser.java

@@ -14 +14 @@
 
 import de.ugoe.cs.eventbench.web.data.WebEvent;
+import de.ugoe.cs.util.console.Console;
 
 public class WeblogParser {
         
         private long timeout;
+        
+        private int minLength = 2;
         
         private List<List<WebEvent>> sequences;
@@ -31 +34 @@
         public List<List<WebEvent>> getSequences() {
                 return sequences; 
+        }
+        
+        public void setTimeout(long timeout) {
+                this.timeout = timeout;
+        }
+        
+        public void setMinLength(int minLength) {
+                this.minLength = minLength;
         }
         
@@ -89 +100 @@
                         }
                 }
+                Console.traceln(""+sequences.size()+ " user sequences found");
+                // prune sequences shorter than min-length
+                for( int i=0; i<sequences.size(); i++ ) {
+                        if( sequences.get(i).size()<minLength ) {
+                                sequences.remove(i);
+                        }
+                }
+                Console.traceln(""+sequences.size()+ " remaining after pruning of sequences shorter than " + minLength);
         }
 }