WUME Lab Projects

Below we list many of the projects underway in the WUME Lab. If you are a student and would like to participate in one of these, please contact Prof. Davison.

Contextual Link Analysis. In this NSF-funded project we consider the context in which a link is found. Such contexts can be incorporated into the link analysis to improve ranking quality.

Understanding and Enhancing Queries. This NSF-funded project considers topic distributions in queries and results, and exploits the query and results bipartite graph to provide related queries, related pages, and suggested queries for websites.

Search Engine Spam Identification and Elimination. Search engine spam reduces the quality of search results. By analyzing page contents and the link graph, we can identify particular spamming techniques and reduce their effect on link analysis algorithms.

Completed projects include:

EDIFY Disruption and Delay Tolerant Networking. In this DARPA-supported project, we design a store-and-forward messaging architecture that sits on top of an existing network infrastructure that is tolerant of delays, disruptions, and node mobility.

Mozilla Prefetching. We have implemented history and content-based prefetching systems and have added them to the open source Mozilla browser.

DNSflow. DNSflow captures DNS packets and logs contents and timestamps for offline analysis.

HTTPflow. Based on tcpflow, HTTPflow captures port 80 (Web) traffic and writes the headers and timestamps to a log for offline analysis. Using this tool we can capture much more detail (especially with respect to caching) than is recorded by Web servers or proxy caches. It also does not require any changes to the clients or servers.

LEAP: Lehigh Evaluation Architecture for Proxies. In this project, we implemented a Simultaneous Proxy Evaluation (SPE) prototype.

Apache Prefetching Hints. We have modified Apache to generate HTTP prefetching hints as supported by the open source Mozilla and Firefox browsers.

Gnutella Network Analysis. This research focuses primarily on measuring, documenting and understanding the network, which should allow for future improvements to the network architecture, and to extensions of the network functionality.