We present SETS, an architecture for building topic-segmented networks for efficient search. The key idea is to arrange participants in a topic-segmented topology where most of the links are short-distance links joining pairs of sites with similar content. The resulting topically focused regions are joined together into a single network by long-distance links. Queries are then matched and routed to only the topically closest regions. We draw on ideas from machine learning and social network theory to build an efficient search network. We discuss a variety of design issues and tradeoffs that an implementor of SETS would face. We show that SETS is ecient in network traffic and query processing load.

Collection selection has been a research issue for years. Typically, in related work, precomputed statistics are employed in order to estimate the expected result quality of each collection, and subsequently the collections are ranked accordingly. Our thesis is that this simple approach is insufficient for several applications in which the collections typically overlap. This is the case, for example, for the collections built by autonomous peers crawling the web. We argue for the extension of existing quality measures using estimators of mutual overlap among collections and present experiments in which this combination outperforms CORI, a popular approach based on quality estimation. We outline our prototype implementation of a P2P web search engine, coined MINERVA 1, that allows handling large amounts of data in a distributed and self-organizing manner. We conduct experiments which show that taking overlap into account during collection selection can drastically decrease the number of collections that have to be contacted in order to reach a satisfactory level of recall, which is a great step toward the feasibility of distributed web search.

This article presents a semisupervised learning solution to the result merging problem. The key contribution is the observation that information used to create resource descriptions for resource selection can also be used to create a centralized sample database to guide the normalization of document scores returned by different databases. At retrieval time, the query is sent to the selected databases, which return database-specific document scores, and to a centralized sample database, which returns database-independent document scores. Documents that have both a database-specific score and a database-independent score serve as training data for learning to normalize the scores of other documents. An extensive set of experiments demonstrates that this method is more effective than the well-known CORI result-merging algorithm under a variety of conditions

Database selection is an important step when searching over large numbers of distributed text databases. The database selection task relies on statistical summaries of the database contents, which are not typically exported by databases. Previous research has developed algorithms for constructing an approximate content summary of a text database from a small document sample extracted via querying. Unfortunately, Zipf’s law practically guarantees that content summaries built this way for any relatively large database will fail to cover many low-frequency words. Incomplete content summaries might negatively affect the database selection process, especially for short queries with infrequent words. To improve the coverage of approximate content summaries, we build on the observation that topically similar databases tend to have related vocabularies. Therefore, the approximate

Peer-to-Peer (P2P) search requires intelligent decisions for query routing: selecting the best peers to which a given query, initiated at some peer, should be forwarded for retrieving additional search results. These decisions are based on statistical summaries for each peer, which are usually organized on a per-keyword basis and managed in a distributed directory of routing indices. Such architectures disregard the

Abstract. Collection selection is one of the key problems in distributed information retrieval. Due to resource constraints it is not usually feasible to search all collections in response to a query. Therefore, the central component (broker) selects a limited number of collections to be searched for the submitted queries. During the past decade, several collection selection algorithms have been introduced. However, their performance varies on different testbeds. We propose a new collection-selection method based on the ranking of downloaded sample documents. We test our method on six testbeds and show that our technique can significantly outperform other state-of-the-art algorithms in most cases. We also introduce a new testbed based on the trec gov2 documents. 1

The TREC .GOV collection makes a valuable web testbed for distributed information retrieval methods because it is naturally partitioned and includes 725 web-oriented queries with judged answers. It can usefully model aspects of government and large corporate portals. Analysis of the .gov data shows that a purely distributed approach would not be feasible for providing search on a .gov portal because of the large number (17,000+) of web sites and the high proportion that do not provide a search interface. An alternative hybrid approach, combining both distributed and centralized techniques, is proposed and server selection methods are evaluated within this framework using web-oriented evaluation methodology. A number of well-known algorithms are compared against representatives (highest anchor ranked page (HARP) and anchor weighted sum (AWSUM)) of a family of new selection methods which use link anchortext extracted from an auxiliary crawl to provide descriptions of sites which are not themselves crawled. Of the previously published methods, ReDDE substantially outperformed three variants of CORI and also outperformed a method based on Kullback-Leibler Divergence (extended) except on topic distillation. HARP and AWSUM performed best overall but were outperformed on the topic distillation task by extended KL Divergence.

Abstract: This paper presents the MINERVA project that protoypes a distributed search engine based on P2P techniques. MINERVA is layered on top of a Chord-style overlay network and uses a powerful crawling, indexing, and search engine on every autonomous peer. We formalize our system model and identify the problem of efficiently selecting promising peers for a query as a pivotal issue. We revisit existing approaches to the database selection problem and adapt them to our system environment. Measurements are performed to compare different selection strategies using real-world data. The experiments show significant performance differences between the strategies and prove the importance of a judicious peer selection strategy. The experiments also present first evidence that a small number of carefully selected peers already provide the vast majority of all relevant results. 1

Abstract. We consider a collaboration of peers autonomously crawling the Web. A pivotal issue when designing a peer-to-peer (P2P) Web search engine in this environment is query routing: selecting a small subset of (a potentially very large number of relevant) peers to contact to satisfy a keyword query. Existing approaches for query routing work well on disjoint data sets. However, naturally, the peers ’ data collections often highly overlap, as popular documents are highly crawled. Techniques for estimating the cardinality of the overlap between sets, designed for and incorporated into information retrieval engines are very much lacking. In this paper we present a comprehensive evaluation of appropriate overlap estimators, showing how they can be incorporated into an efficient, iterative approach to query routing, coined Integrated Quality Novelty (IQN). We propose to further enhance our approach using histograms, combining overlap estimation with the available score/ranking information. Finally, we conduct a performance evaluation in MINERVA, our prototype P2P Web search engine.

Exploiting information induced from (query-specific) clustering of top-retrieved documents has long been proposed as means for improving precision at the very top ranks of the returned results. We present a novel language model approach to ranking query-specific clusters by the presumed percentage of relevant documents that they contain. While most previous cluster ranking approaches focus on the cluster as a whole, our model also exploits information induced from documents associated with the cluster. Our model substantially outperforms previous approaches for identifying clusters containing a high relevant-document percentage. Furthermore, using the model to produce document ranking yields precision-at-top-ranks performance that is consistently better than that of the initial ranking upon which clustering is performed; the performance also favorably compares with that of a state-of-the-art pseudo-feedback retrieval method.