Hardware Accelerated Algorithms for Semantic Processing of Document Streams

Eick, Stephen G.; Lockwood, John W.; Loui, Ronald P.; Levine, Andrew; Mauger, Justin; Weishar, D.; Ratner, Alan M.; Byrnes, John

doi:10.1109/aero.2006.1656050

Cited by 17 publications

(7 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It only requires memory to store the current solution. This together with the simplicity of the method makes it suitable for implementation in specialized hardware [9]. The method described here uses the simplistic maximum likelihood method for estimating probability distributions from number counts.…”

Section: Discussionmentioning

confidence: 99%

Simmered Greedy Optimization for Co-clustering

Kapadia

Rohwer

2010

2010 Seventh International Conference on Information Technology: New Generations

View full text Add to dashboard Cite

We present a fast yet highly effective stochastic algorithm, Simmered Greedy Optimization (SG(N)) for solving the coclustering problem: to simultaneously cluster two finite sets by maximizing the mutual information between the clusterings. (Clustering one set by this criterion is a special case.) This is a combinatorial optimization problem of great interest for deriving maximally predictive feature sets. Co-clustering has found applications in many areas, particularly statistical natural language processing and bioinformatics. We report results of tests on a suite of statistical natural language problems, comparing SG(N) with simulated annealing and a publicly available implementation of co-clustering. In all cases we obtain superior results with far less computation using SG(N).

show abstract

Section: Discussionmentioning

confidence: 99%

Simmered Greedy Optimization for Co-clustering

Kapadia

Rohwer

2010

2010 Seventh International Conference on Information Technology: New Generations

View full text Add to dashboard Cite

show abstract

“…Overall, our application performs at a much higher I/O rate (800MB/s per FPGA, 3.2GB/s for the board) than required by Netezza. Other FPGA systems have also been studied for Information Retrieval, specifically for mapping Language Classification algorithms using Ngrams [20], [21]. Lockwood [20] implemented a Language Classification algorithm to differentiate 255 languages.…”

Section: Related Workmentioning

confidence: 99%

“…Other FPGA systems have also been studied for Information Retrieval, specifically for mapping Language Classification algorithms using Ngrams [20], [21]. Lockwood [20] implemented a Language Classification algorithm to differentiate 255 languages. Since the embedded memory available on FPGAs is limited, they mapped the Bloom Filter on external memory resulting in lower throughputs due to the random memory access.…”

Section: Related Workmentioning

confidence: 99%

Evaluating FPGA-acceleration for real-time unstructured search

Chalamalasetti

Margala

Vanderbauwhede

et al. 2012

2012 IEEE International Symposium on Performance Analysis of Systems &Amp; Software

View full text Add to dashboard Cite

Abstract-Emerging data-centric workloads that operate on and harvest useful insights from large amounts of unstructured data require corresponding new data-centric system architecture optimizations. In particular, with the growing importance of power and cooling costs, a key challenge for such future designs is to achieve increased performance at high energy efficiency. At the same time, recent trends towards better support for reconfigurable logic enable the use of energy-efficient accelerators. Combining these trends, in this paper, we examine the applicability of acceleration in future data-centric system architectures. We focus on an important class of data-centric workloads, realtime unstructured search, or information filtering, where large collections of documents are scored against specific topic profiles, and present an FPGA-based implementation to accelerate such workloads. Our implementation, based on the GiDEL PROCStar IV board using Altera Stratix IV FPGAs, demonstrates excellent performance and energy efficiency, 20 to 40 times better than baseline server systems for typical usage scenarios. Our results also highlight interesting insights for the design of accelerators in future data-centric systems.

show abstract

“…We used a subset of the CMU 20 Newsgroups data set containing approximately 13,000 postings from 13 different newsgroups [3], [14]. The full vocabulary for this data set consists of approximately 117,000 terms.…”

Section: Resultsmentioning

confidence: 99%

“…We report a new development in an ongoing research project [1], [3], [9], [10] targeted at analyzing massive amounts of network data, both as archived data sets and as network traffic streaming at rates up to and even exceeding 2.4Gbps.…”

Section: Introductionmentioning

confidence: 99%

An Architecture for Streaming Coclustering in High Speed Hardware

Byrnes

Rohwer

2006 IEEE Aerospace Conference

Self Cite

View full text Add to dashboard Cite

We seek to learn the semantics of a data stream at optical line speed. 1,2 We focus on text data, but the techniques developed should apply to broad modalities of network data wherever appropriate features can be computed rapidly enough. We consider a custom hardware system designed to categorize documents based on feature clusters and document clusters that have been learned offline on standard general-purpose computers, and we present a technique for extending this system to permit online learning from arbitrarily large data sets.

show abstract

Hardware Accelerated Algorithms for Semantic Processing of Document Streams

Cited by 17 publications

References 15 publications

Simmered Greedy Optimization for Co-clustering

Simmered Greedy Optimization for Co-clustering

Evaluating FPGA-acceleration for real-time unstructured search

An Architecture for Streaming Coclustering in High Speed Hardware

Contact Info

Product

Resources

About