Peer network systems are becoming an increasingly important development in Web search technology. Many studies show that peer search systems perform better when a query is sent to a group of peers semantically similar to the query. This suggests that semantic communities should form so that a query can quickly propagate to many appropriate peers. For the network to be functional, its dynamic communication topology must match the semantic clustering of peers. We introduce two criteria to evaluate a peer search network based on the concept of semantic locality: first, the "smallworld" topology of the network; second, we use topical semantic similarity to monitor the quality of a peer's neighbors over time by looking at whether a peer chooses semantically appropriate neighbors to route its queries. We present several simulation experiments conducted with different peer search algorithms on our peer Web search system, 6S. The results suggest that 6S, despite its use of an unstructured overlay network; can effectively foster the spontaneous formation of semantic communities through local peer interactions alone.

ExpertEyes is a low-cost, open-source package of hardware and software that is designed to provide portable high-definition eyetracking. The project involves several technological innovations, including portability, high-definition video recording, and multiplatform software support. It was designed for challenging recording environments, and all processing is done offline to allow for optimization of parameter estimation. The pupil and corneal reflection are estimated using a novel forward eye model that simultaneously fits both the pupil and the corneal reflection with full ellipses, addressing a common situation in which the corneal reflection sits at the edge of the pupil and therefore breaks the contour of the ellipse. The accuracy and precision of the system are comparable to or better than what is available in commercial eyetracking systems, with a typical accuracy of less than 0.4° and best accuracy below 0.3°, and with a typical precision (SD method) around 0.3° and best precision below 0.2°. Part of the success of the system comes from a high-resolution eye image. The high image quality results from uncasing common digital camcorders and recording directly to SD cards, which avoids the limitations of the analog NTSC format. The software is freely downloadable, and complete hardware plans are available, along with sources for custom parts.

Sixearch.org is a peer application for social, distributed, adaptive Web search, which integrates the Sixearch.org protocol, a topical crawler, a document indexing system, a retrieval engine, a P2P network communication system, and a contextual learning system. With a single click, the Sixearch.org application will build your personal Web collection. You can search not only your collection, but also other Sixearch peers. When you submit a query, your Sixearch agent will determine which peers are best suited to answer it based on previous interactions. Your agent will also learn from the results it receives, so that it can continuously improve.

No abstract

C entralized search engines cannot cover the entire web (Lawrence and Giles 1999) because it is too large, fast-growing and fast-changing (Brewington and Cybenko 2000;Fetterly et al. 2003;Ntoulas, Cho, and Olston 2004). As a result, current centralized search engines focus on "important" portions of the web. However, the notion of importance is highly subjective: the biases that are introduced to address the needs of the "average" user can result in diminished effectiveness in satisfying many atypical search needs. Therefore, the "one engine fits all" model cannot handle the increasing size, rate of change, and heterogeneity of the web and its users. In addition, as search becomes more prevalent at the desktop level, users will increasingly want to make subsets of the files indexed in their computers available to others through the Internet. Peer networks provide us with an architecture for extending web search technology to capture the contextual needs of a diverse population of users, while leveraging their resources.There are several models of peer network topologies and query protocols, including structured, unstructured, flooding, distributed hash tables, and hierarchical (Androutsellis-Theotokis and Spinellis 2004). Our design of a collaborative web search network is guided by the principle of semantic locality: peers with shared interests are likely to communicate with each other more frequently than unrelated agents, so they should be able to reach each other in a few virtual hops. However, a dense network would generate too much traffic. A good topology favors both effectiveness and efficiency, by making it possible for a query to reach a relevant target peer in few steps, without imposing a large traffic load on the entire network. Small-world networks (Watts and Strogatz 1998) provide both clustered communities and enough randomness to keep the network distance small between any two peers. Effective search requires that the clusters be associated with a high semantic similarity between neighbors (Watts, Dodds, and Newman 2002). Because there is no global knowledge of the network (what peers are currently present, what information they hold, and what information they seek), and the network is very dynamic (peers may join and leave the network at any time), we cannot impose semantic locality into the network by design; instead, we explore AI techniques through which semantic locality will emerge as the result of local interactions and learning by individual peer agents.Our research group is currently developing 6S, an intelligent multiagent application for peer-based web search (Wu, Akavipat, and Menczer 2005;Akavipat et al. 2006). The name is a contraction of "six degrees of separation" and "search," to reflect the social network of peer agents at the base of the collaborative search process. Each 6S peer agent is both a (limited) directory hub and a content provider; it has its own topical crawler (based on local context), which supports a