Discussion Tracking in Enron Email Using PARAFAC

Bader, Brett W.; Berry, Michael W.; Browne, Murray

doi:10.1007/978-1-84800-046-9_8

Cited by 81 publications

(73 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Also CP decomposition has been used for noise removal and data compression [24]. In addition, it has a wide range of applications in many scientific areas such as Data Mining [8] and telecommunications [34]. There are few other factorization methods which mainly originate from CP and/or Tucker methods.…”

Section: )mentioning

confidence: 99%

A tensor-based selection hyper-heuristic for cross-domain heuristic search

Asta

Özcan

2015

Information Sciences

View full text Add to dashboard Cite

Hyper-heuristics have emerged as automated high level search methodologies that manage a set of low level heuristics for solving computationally hard problems. A generic selection hyper-heuristic combines heuristic selection and move acceptance methods under an iterative single point-based search framework. At each step, the solution in hand is modified after applying a selected heuristic and a decision is made whether the new solution is accepted or not. In this study, we represent the trail of a hyper-heuristic as a third order tensor. Factorization of such a tensor reveals the latent relationships between the low level heuristics and the hyper-heuristic itself. The proposed learning approach partitions the set of low level heuristics into two equal subsets where heuristics in each subset are associated with a separate move acceptance method. Then a multi-stage hyper-heuristic is formed and while solving a given problem instance, heuristics are allowed to operate only in conjunction with the associated acceptance method at each stage. To the best of our knowledge, this is the first time tensor analysis of the space of heuristics is used as a data science approach to improve the performance of a hyper-heuristic in the prescribed manner. The empirical results across six different problem domains from a benchmark indeed indicate the success of the proposed approach.

show abstract

Section: )mentioning

confidence: 99%

A tensor-based selection hyper-heuristic for cross-domain heuristic search

Asta

Özcan

2015

Information Sciences

View full text Add to dashboard Cite

show abstract

“…The first data set include dense tensors using synthetically generated factors. For each of K = 10, 20, 60, and 120, we constructed A (1) , A (2) , and A (3) of size 300 × K using random numbers from the uniform distribution over [0,1]. Then, we randomly selected 50 percent of elements in A (1) , A (2) , and A (3) to make them zero.…”

Section: Data Setsmentioning

confidence: 99%

“…For each of K = 10, 20, 60, and 120, we constructed A (1) , A (2) , and A (3) of size 300 × K using random numbers from the uniform distribution over [0,1]. Then, we randomly selected 50 percent of elements in A (1) , A (2) , and A (3) to make them zero. Finally, a three way tensor of size 300 × 300 × 300 is constructed by A (1) , A (2) , A (3) .…”

Section: Data Setsmentioning

confidence: 99%

“…Then, we randomly selected 50 percent of elements in A (1) , A (2) , and A (3) to make them zero. Finally, a three way tensor of size 300 × 300 × 300 is constructed by A (1) , A (2) , A (3) . Different tensors were created for different K values.…”

Section: Data Setsmentioning

confidence: 99%

“…Interestingly, some method for finding nonnegative factors of higher-order tensors, such as [6], were introduced even before NMF. Recent work dealt with properties such as degeneracy [23] and applications such as sound source separation [9], text mining [2], and computer vision [27].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Fast Nonnegative Tensor Factorization with an Active-Set-Like Method

Kim

Park

2012

High-Performance Scientific Computing

View full text Add to dashboard Cite

We introduce an efficient algorithm for computing a low-rank nonnegative CANDECOMP/PARAFAC (NNCP) decomposition. In text mining, signal processing, and computer vision among other areas, imposing nonnegativity constraints to the low-rank factors of matrices and tensors has been shown an effective technique providing physically meaningful interpretation. A principled methodology for computing NNCP is alternating nonnegative least squares, in which the nonnegativity-constrained least squares (NNLS) problems are solved in each iteration. In this chapter, we propose to solve the NNLS problems using the block principal pivoting method. The block principal pivoting method overcomes some difficulties of the classical active method for the NNLS problems with a large number of variables. We introduce techniques to accelerate the block principal pivoting method for multiple right-hand sides, which is typical in NNCP computation. Computational experiments show the state-of-the-art performance of the proposed method.

show abstract