Skip graphs are a novel distributed data structure, based on skip lists, that provide the full functionality of a balanced tree in a distributed system where resources are stored in separate nodes that may fail at any time. They are designed for use in searching peer-to-peer systems, and by providing the ability to perform queries based on key ordering, they improve on existing search tools that provide only hash table functionality. Unlike skip lists or other tree data structures, skip graphs are highly resilient, tolerating a large fraction of failed nodes without losing connectivity. In addition, simple and straightforward algorithms can be used to construct a skip graph, insert new nodes into it, search it, and detect and repair errors within it introduced due to node failures.

In this paper, we explore the use of domain-independent and domain-specific ontologies to find matching service descriptions. The domain-independent relationships are derived using an English thesaurus after tokenization and part-of-speech tagging. The domain-specific ontological similarity is derived by an inference on the semantic annotations associated with Web service descriptions. Matches due to the two cues are combined to determine an overall semantic similarity score. By combining multiple cues, we show that better relevancy results can be obtained for service matches from a large repository, than could be obtained using any one cue alone.

In this paper, we address the problem of searching schema databases for semantically-related schemas. We first give a method of finding semantic similarity between pair-wise schemas based on tokenization, part-of-speech tagging, word expansion, and ontology matching. We then address the problem of indexing the schema database through a semantic hash table. Matching schemas in the database are found by hashing the query attributes and recording peaks in the histogram of schema hits. Results indicated a 90% improvement in search performance while maintaining high precision and recall.

We consider the problem of designing an overlay network and routing mechanism that permits finding resources efficiently in a peer-to-peer system. We argue that many existing approaches to this problem can be modeled as the construction of a random graph embedded in a metric space whose points represent resource identifiers, where the probability of a connection between two nodes depends only on the distance between them in the metric space. We study the performance of a peer-to-peer system where nodes are embedded at grid points in a simple metric space: a one-dimensional real line. We prove upper and lower bounds on the message complexity of locating particular resources in such a system, under a variety of assumptions about failures of either nodes or the connections between them. Our lower bounds in particular show that the use of inverse power-law distributions in routing, as suggested by Kleinberg [5], is close to optimal. We also give heuristics to efficiently maintain a network supporting efficient routing as nodes enter and leave the system. Finally, we give some experimental results that suggest promising directions for future work.

A randomized algorithm is given that solves the wait-free consensus problem for a shared-memory model with infinitely many processes. The algorithm is based on a weak shared coin algorithm that uses weighted voting to achieve a majority outcome with at least constant probability that cannot be disguised even if a strong adversary is allowed to destroy infinitely many votes. The number of operations performed by process i is a polynomial function of i. Additional algorithms are given for solving consensus more efficiently in models with an unknown upper bound b on concurrency or an unknown upper bound n on the number of active processes; under either of these restrictions, it is also shown that the problem can be solved even with infinitely many anonymous processes by prefixing each instance of the shared coin with a naming algorithm that breaks symmetry with high probability. For many of these algorithms, matching lower bounds are proved that show that their per-process work is nearly optimal as a function of i, b, or n. The case of n active processes gives an algorithm for anonymous, adaptive consensus that requires only O(n log2 n) per-process work, which is within a constant factor of the best previously known non-adaptive algorithm for a strong adversary. Finally, it is shown that standard universal constructions based on consensus continue to work with infinitely many processes with only slight modifications. This shows that in infinite distributed systems, as in finite ones, with randomness all things are possible.