On an Algorithm for Identifying Sessions from Web Logs

The quality of decisions is based on the quality of processed data. So it is important that at the beginning of the data mining process to provide correct and quality data. The preprocessing data is a necessity for avoiding the failure of the data analysis. The idea that the data mining process can be done without human supervision has proved to be wrong. Even so, the humans are trying to automate as much as possible the process. From here are resulting many algorithms and techniques that are implemented using various programming language. In this work is presented an algorithm for identifying the sessions from a web logs file. It uses a value of 30 minutes to mark the end of a session and start another. We compute the average time for visiting the pages and using this we show that the presented algorithm produces errors in identifying sessions. We consider that the correct way to identify the session is to take into account the average time for visiting the pages.