Identifying User Actions from Network Traffic
Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Identification of a user’s actions while browsing the Internet is mostly achieved by instrumentation of the user’s browser or by obtaining server logs. In both cases this requires installation of software on multiple clients and/or servers in order to obtain sufficient data. However, by using network traffic, access to user generated traffic from multiple clients to multiple servers is possible. In this project a proxy server is used for recording network traffic and a user-action identification algorithm is proposed. The proposed algorithm includes various policies of analyzing network traffic in order to identify user actions. This project also presents an evaluation framework for the proposed policies, based on which the tradeoff of the various policies is revealed. Proxy servers are widely deployed by numerous organizations and often used for web mining, so with the work of this project user action recognition can be a new tool when considering web traffic evaluation.
Place, publisher, year, edition, pages
2015. , 48 p.
proxy logs, network traffic
IdentifiersURN: urn:nbn:se:liu:diva-119675ISRN: LIU-IDA/LITH-EX-A--15/036--SEOAI: oai:DiVA.org:liu-119675DiVA: diva2:825810
Subject / course
2015-06-08, Muhammad al-Khwarizmi, 15:00 (English)
Krishnamoorthi, Vengatanathan, PhD student
Carlsson, Niklas, Associate Professor