Abstract: Counting in general, and estimating the cardinality of (multi-) sets in particular, is highly desirable for a large variety of applications, representing a foundational block for the efficient deployment and access of emerging internet-scale information systems. Examples of such applications range from optimizing query access plans in internet-scale databases, to evaluating the significance (rank/score) of various data items in informationretrieval applications. The key constraints that any acceptable solution must satisfy are: (i) efficiency: the number of nodes that need be contacted for counting purposes must be small in order to enjoy small latency and bandwidth requirements; (ii) scalability, seemingly contradicting the efficiency goal: arbitrarily large numbers of nodes nay need to add elements to a (multi-) set, which dictates the need for a highly distributed solution, avoiding server-based scalability, bottleneck, and availability problems; (iii) access and storage load balancing: counting and related overhead chores should be distributed fairly to the nodes of the network; (iv) accuracy: tunable, robust (in the presence of dynamics and failures) and highly accurate cardinality estimation; (v) simplicity and ease of integration: special, solution-specific indexing structures should be avoided. In this paper, first we contribute a highly-distributed, scalable, efficient, and accurate (multi-) set cardinality estimator. Subsequently, we show how to use our solution to build and maintain histograms, which have been a basic building block for query optimization for centralized databases, facilitating their porting into the realm of internet-scale data networks.
Abstract: Counting items in a distributed system, and estimating the cardinality of multisets in particular,
is important for a large variety of applications and a fundamental building block for emerging Internet-scale information systems. Examples of such applications range from optimizing query access plans in peer-to-peer data sharing, to computing the significance (rank/score) of data items in distributed informationretrieval. The general formal problem addressed in this article is computing the network-wide distinct number of items with some property (e.g., distinct files with file name
containing “spiderman”) where each node in the network holds an arbitrary subset, possibly overlapping the subsets of other nodes. The key requirements that a viable approach must satisfy are:
(1) scalability towards very large network size, (2) efficiency regarding messaging overhead, (3) load
balance of storage and access, (4) accuracy of the cardinality estimation, and (5) simplicity and easy
integration in applications. This article contributes the DHS (Distributed Hash Sketches) method
for this problem setting: a distributed, scalable, efficient, and accurate multiset cardinality estimator.
DHSis based on hash sketches for probabilistic counting, but distributes the bits of each counter
across network nodes in a judicious manner based on principles of Distributed Hash Tables, paying
careful attention to fast access and aggregation as well as update costs. The article discusses various
design choices, exhibiting tunable trade-offs between estimation accuracy, hop-count efficiency, and
load distribution fairness. We further contribute a full-fledged, publicly available, open-source implementation of all our methods, and a comprehensive experimental evaluation for various settings.
Abstract: Informationretrieval (IR) in peer-to-peer (P2P) networks,
where the corpus is spread across many loosely coupled
peers, has recently gained importance. In contrast to IR
systems on a centralized server or server farm, P2P IR faces
the additional challenge of either being oblivious to global
corpus statistics or having to compute the global measures
from local statistics at the individual peers in an efficient,
distributed manner. One specific measure of interest is the
global document frequency for different terms, which would
be very beneficial as term-specific weights in the scoring and
ranking of merged search results that have been obtained
from different peers.
This paper presents an efficient solution for the problem
of estimating global document frequencies in a large-scale
P2P network with very high dynamics where peers can join
and leave the network on short notice. In particular, the
developed method takes into account the fact that the lo-
cal document collections of autonomous peers may arbitrar-
ily overlap, so that global counting needs to be duplicate-
insensitive. The method is based on hash sketches as a
technique for compact data synopses. Experimental stud-
ies demonstrate the estimator?s accuracy, scalability, and
ability to cope with high dynamics. Moreover, the benefit
for ranking P2P search results is shown by experiments with
real-world Web data and queries.
Abstract: Informationretrieval (IR) in peer-to-peer (P2P) networks,
where the corpus is spread across many loosely coupled
peers, has recently gained importance. In contrast to IR
systems on a centralized server or server farm, P2P IR faces
the additional challenge of either being oblivious to global
corpus statistics or having to compute the global measures
from local statistics at the individual peers in an efficient,
distributed manner. One specific measure of interest is the
global document frequency for different terms, which would
be very beneficial as term-specific weights in the scoring and
ranking of merged search results that have been obtained
from different peers.
This paper presents an efficient solution for the problem
of estimating global document frequencies in a large-scale
P2P network with very high dynamics where peers can join
and leave the network on short notice. In particular, the
developed method takes into account the fact that the lo-
cal document collections of autonomous peers may arbitrar-
ily overlap, so that global counting needs to be duplicate-
insensitive. The method is based on hash sketches as a
technique for compact data synopses. Experimental stud-
ies demonstrate the estimator?s accuracy, scalability, and
ability to cope with high dynamics. Moreover, the benefit
for ranking P2P search results is shown by experiments with
real-world Web data and queries.
Abstract: The peer-to-peer computing paradigm is an intriguing alternative to Google-style search
engines for querying and ranking Web content. In a network with many thousands or
millions of peers the storage and access load requirements per peer are much lighter
than for a centralized Google-like server farm; thus more powerful techniques from informationretrieval, statistical learning, computational linguistics, and ontological reasoning
can be employed on each peer¢s local search engine for boosting the quality
of search results. In addition, peers can dynamically collaborate on advanced and particularly
difficult queries. Moroever, a peer-to-peer setting is ideally suited to capture
local user behavior, like query logs and click streams, and disseminate and aggregate
this information in the network, at the discretion of the corresponding user, in order to
incorporate richer cognitive models.
This paper gives an overview of ongoing work in the EU Integrated Project DELIS
that aims to develop foundations for a peer-to-peer search engine with Google-or-better
scale, functionality, and quality, which will operate in a completely decentralized and
self-organizing manner. The paper presents the architecture of such a system and the
Minerva prototype testbed, and it discusses various core pieces of the approach: efficient
execution of top-k ranking queries, strategies for query routing when a search request
needs to be forwarded to other peers, maintaining a self-organizing semantic overlay
network, and exploiting and coping with user and community behavior.