Reproduction of all or part of this work is permitted for educational or research use on condition that this copyright notice is included in any copy. Publications in the FI MU Report Series are in general accessible via WWW:

Effective peer selection for intelligent query routing is a challenge in collaborative peer-based Web search systems, especially unstructured networks that do not have any centralized control of peer document collections. In particular, routing a query to multiple peers that provide the same results is a waste of resources. To deal with overlapping document collections we propose a diverse peer selection approach for adaptive query routing. This approach takes into account not only which neighbors are the best resource providers for a given query, but also which combinations of neighbors can provide the least redundant results. We validate the feasibility of our proposed algorithm by presenting several simulation experiments conducted with different configurations of peer network environments. Two novel evaluation measures, distributed precision and distributed recall, are also introduced to provide an effective comparison of different peer network systems. These two performance measures extend the well known IR measures of precision and recall by integrating network costs, namely bandwidth and latency. Our algorithm finds results of equivalent quality using less time and generating less traffic in the presence of varying amounts of document duplication.

Abstract This report motivates the need for large-scale distributed approaches to information retrieval, and proposes

Collaborative query routing is a new paradigm for Web search that treats both established search engines and other publicly available indices as intelligent peer agents in a search network. The approach makes it transparent for anyone to build their own (micro) search engine, by integrating established Web search services, desktop search, and topical crawling techniques. The challenge in this model is that each of these agents must learn about its environment — the existence, knowledge, diversity, reliability, and trustworthiness of other agents — by analyzing the queries received from and results exchanged with these other agents. We present the 6S peer network, which uses machine learning techniques to learn about the changing query environment. We show that simple reinforcement learning algorithms are sufficient to detect and exploit semantic locality in the network, resulting in efficient routing and high-quality search results. A prototype of 6S is available for public use and is intended to assist in the evaluation of different AI techniques employed by the networked agents.

Despite the many research efforts invested recently in peerto-peer search engines, none of the proposed system has reached the level of quality and efficiency of their centralized counterpart. One of the main reasons for this inferior performance is the difficulty to attract a critical mass of users that would make the peer-to-peer system truly competitive. We argue that decentralized search mechanisms should not aim at replacing existing engines, but instead complement them by adding novel functionalities that would be difficult to provide in a centralized manner. This paper introduces an example of such a complementary search mechanism and presents the design of a distributed collaborative system for leveraging user feedback and document/user profiling information.

6S is a collaborative peer network application, aimed to extend the current model of centralized search engines with large numbers of autonomous, distributed, micro-search engines. Each peer within the 6S network crawls the Web in a focused way, guided by its user’s information context. This way better contextual coverage can be achieved. Each peer also acts within the network by submitting, forwarding, and responding to queries to/from its neighbors. Peers depend on a local adaptive routing algorithm to dynamically change the topology of the peer network and search for the best neighbors to answer their queries. If realized, the 6S network will lead to an “intelligent network” where peers learn continually about users and other peers, thus addressing user search needs in a contextual, personalized, and scalable way. Experimental plan The initial phase of this project was funded by a NSF Career grant. This has proven the viability of the 6S idea via large-scale simulation, and has led to the development of a first working prototype. At the current stage, each peer is identical in functionality and differs only in the focus of its collection; all peers employ the same keyword interface to retrieve and rank results, and the same learning algorithm to track other peers and route queries. Now we propose to deploy such application in the real world and allow for the development of peers with different focus,

Abstract. This paper is concerned with the evaluation of distributed and peer-to-peer information retrieval systems. A new measure is introduced that compares results of a distributed retrieval system to those of a centralised system, fully exploiting the ranking of the latter as an indicator of gradual relevance. Problems with existing evaluation approaches are verified experimentally. 1

For most users, Web-based centralized search engines are the access point to distributed resources such as Web pages, items shared in file sharing-systems, etc. Unfortunately, existing search engines compute their results on the basis of structural information only, e.g., the Web graph structure or query-document similarity estimations. Users expectations are rarely considered to enhance the subjective relevance of returned results. However, exploiting such information can help search engines satisfy users by tailoring search results. Interestingly, user interests typically follow the clustering property: users who were interested in the same topics in the past are likely to be interested in these same topics also in the future. It follows that search results considered relevant by a user belonging to a group of homogeneous users will likely also be of interest to other users from the same group. In this paper, we propose the architecture of a novel peerto-peer system exploiting collaboratively built search mechanisms. The paper discusses the challenges associated with a system based on the interest clustering principle. The objective is to provide a self-organized network of users, grouped according to the interests they share, that can be leveraged to enhance the quality of the experience perceived by users searching the Web.

Proceedings Copyright c ○ 2009 for the individual papers by the papers ’ authors. Copying permitted for private and academic purposes. Re-publication of material from this volume requires permission by the copyright owners.