Evaluation of retrieval performance is a crucial problem in content-based image retrieval CBIR. Many di erent methods for measuring the performance of a system have been created and used by researchers. This article discusses the advantages and shortcomings of the performance measures currently used. Problems such a s a commonimage database for performance comparisons and a means of getting relevance judgments or ground truth for queries are explained. The relationship between CBIR and information retrieval IR is made clear, since IR researchers have decades of experience with the evaluation problem. Many of their solutions can be used for CBIR, despite the di erences between the elds. Several methods used in text retrieval are explained. Proposals for performance measures and means of developing a standard test suite for CBIR, similar to that used in IR at the annual Text REtrieval Conference TREC, are presented.

Relevance feedback has been shown to be a very effective tool for enhancing retrieval results in text retrieval. In content-based image retrieval it is more and more frequently used and verv good results have been obtained. However, too much negative feedback may destroy a query as good features get negative weightings.This paper compares a variety of strategies for positive and negative feedback. The performance evaluation of feedback algorithms is a hard problem. To solve this, we obtain judgmentsfiom several users and employ an automated feedback scheme. We can then evaluate different techniques using the same judgments. Using automated feedback, the ability of a systeni to adapt to the user's needs can be measured very effectively. Our study highlights the utility of negative feedback, especially over severa1,feedback steps.

The conversion of prion helix 1 from an ␣-helical into an extended conformation is generally assumed to be an essential step in the conversion of the cellular isoform PrP C of the prion protein to the pathogenic isoform PrP Sc . Peptides encompassing helix 1 and flanking sequences were analyzed by nuclear magnetic resonance and circular dichroism. Our results indicate a remarkably high instrinsic helix propensity of the helix 1 region. In particular, these peptides retain significant helicity under a wide range of conditions, such as high salt, pH variation, and presence of organic co-solvents. As evidenced by a data base search, the pattern of charged residues present in helix 1 generally favors helical structures over alternative conformations. Because of its high stability against environmental changes, helix 1 is unlikely to be involved in the initial steps of the pathogenic conformational change. Our results implicate that interconversion of helix 1 is rather representing a barrier than a nucleus for the PrP C 3 PrP Sc conversion.Prion protein, PrP, 1 is probably the disease-causing agent of transmissible spongiform encephalopathies such as bovine spongiform encephalopathy in cattle or Creutzfeldt-Jakob disease in man (1). Its cellular form, PrP C , is a highly conserved cell surface glycoprotein of 230 amino acids expressed in all of the mammals studied so far as well as in several species of fish and birds (2, 3). The physiological function of PrP C is not yet fully understood. PrP C seems to be involved in the maintenance of proper presynaptic copper levels as well as in protecting neurons from oxidative stress (4, 5). In addition, the physiological function of PrP C could be associated with higher neurological functions such as learning and memory (5). According to the protein-only hypothesis, disease is caused by accumulation of a misfolded pathogenic isoform, PrP Sc , which is the result of an irreversible large scale conformational change of PrP C . Although PrP C is largely ␣-helical and soluble in polar solvents and sensitive to protease K digestion, PrP Sc consists mostly of ␤-sheets, is soluble only in nonpolar, denaturing solvents, and is resistant to digestion with protease K (6). PrP Sc forms fibrillar aggregates similar to other amyloid fibrils (7). Accumulation of PrP Sc aggregates is accompanied by astrocytosis and gliosis in central nervous tissue, which in turn result in vacuoles in the brains of patients.The solution structures of human PrP-(23-230), huPrP (8), mouse PrP-(121-231) (9), bovine PrP-(23-230) (10), and Syrian hamster PrP-(29 -231) (11) have been determined by NMR spectroscopy. They possess a high degree of structural conservation consistent with the high sequence identity of these proteins. Prion proteins consist of a flexible NH 2 -terminal domain spanning residues 23-124 (huPrP C -numbering scheme), which is largely disordered. This region includes an octapeptide sequence that is repeated four times from residues 60 to 92 and that is likely to bind copper (4). This part al...

No abstract

Much of the present P2P-IR literature is focused on distributed indexing structures. Within this paper, we present an approach based on the replication of peer data summaries via rumor spreading and multicast in a structured overlay.We will describe Rumorama, a P2P framework for similarity queries inspired by GlOSS and CORI and their P2P-adaptation, PlanetP. Rumorama achieves a hierarchization of PlanetP-like summary-based P2P-IR networks. In a Rumorama network, each peer views the network as a small PlanetP network with connections to peers that see other small PlanetP networks. One important aspect is that each peer can choose the size of the PlanetP network it wants to see according to its local processing power and bandwidth. Even in this adaptive environment, Rumorama manages to process a query such that the summary of each peer is considered exactly once in a network without churn. However, the actual number of peers to be contacted for a query is a small fraction of the total number of peers in the network.Within this article, we present the Rumorama base protocol, as well as experiments demonstrating the scalability and viability of the approach under churn.