Publish in OALib Journal

ISSN: 2333-9721

APC: Only $99


Search Results: 1 - 10 of 2440 matches for " Jordan Boyd-Graber "
All listed articles are free for downloading (OA Articles)
Page 1 /2440
Display every page Item
Syntactic Topic Models
Jordan Boyd-Graber,David M. Blei
Computer Science , 2010,
Abstract: The syntactic topic model (STM) is a Bayesian nonparametric model of language that discovers latent distributions of words (topics) that are both semantically and syntactically coherent. The STM models dependency parsed corpora where sentences are grouped into documents. It assumes that each word is drawn from a latent topic chosen by combining document-level features and the local syntactic context. Each document has a distribution over latent topics, as in topic models, which provides the semantic consistency. Each element in the dependency parse tree also has a distribution over the topics of its children, as in latent-state syntax models, which provides the syntactic consistency. These distributions are convolved so that the topic of each word is likely under both its document and syntactic context. We derive a fast posterior inference algorithm based on variational methods. We report qualitative and quantitative studies on both synthetic data and hand-parsed documents. We show that the STM is a more predictive model of language than current models based only on syntax or only on topics.
Multilingual Topic Models for Unaligned Text
Jordan Boyd-Graber,David Blei
Computer Science , 2012,
Abstract: We develop the multilingual topic model for unaligned text (MuTo), a probabilistic model of text that is designed to analyze corpora composed of documents in two languages. From these documents, MuTo uses stochastic EM to simultaneously discover both a matching between the languages and multilingual latent topics. We demonstrate that MuTo is able to find shared topics on real-world multilingual corpora, successfully pairing related documents across languages. MuTo provides a new framework for creating multilingual topic models without needing carefully curated parallel corpora and allows applications built using the topic model formalism to be applied to a much wider class of corpora.
Using Variational Inference and MapReduce to Scale Topic Modeling
Ke Zhai,Jordan Boyd-Graber,Nima Asadi
Computer Science , 2011,
Abstract: Latent Dirichlet Allocation (LDA) is a popular topic modeling technique for exploring document collections. Because of the increasing prevalence of large datasets, there is a need to improve the scalability of inference of LDA. In this paper, we propose a technique called ~\emph{MapReduce LDA} (Mr. LDA) to accommodate very large corpus collections in the MapReduce framework. In contrast to other techniques to scale inference for LDA, which use Gibbs sampling, we use variational inference. Our solution efficiently distributes computation and is relatively simple to implement. More importantly, this variational implementation, unlike highly tuned and specialized implementations, is easily extensible. We demonstrate two extensions of the model possible with this scalable framework: informed priors to guide topic discovery and modeling topics from a multilingual corpus.
Modeling Images using Transformed Indian Buffet Processes
Ke Zhai,Yuening Hu,Sinead Williamson,Jordan Boyd-Graber
Computer Science , 2012,
Abstract: Latent feature models are attractive for image modeling, since images generally contain multiple objects. However, many latent feature models ignore that objects can appear at different locations or require pre-segmentation of images. While the transformed Indian buffet process (tIBP) provides a method for modeling transformation-invariant features in unsegmented binary images, its current form is inappropriate for real images because of its computational cost and modeling assumptions. We combine the tIBP with likelihoods appropriate for real images and develop an efficient inference, using the cross-correlation between images and features, that is theoretically and empirically faster than existing inference techniques. Our method discovers reasonable components and achieve effective image reconstruction in natural images.
Linguistic Harbingers of Betrayal: A Case Study on an Online Strategy Game
Vlad Niculae,Srijan Kumar,Jordan Boyd-Graber,Cristian Danescu-Niculescu-Mizil
Computer Science , 2015,
Abstract: Interpersonal relations are fickle, with close friendships often dissolving into enmity. In this work, we explore linguistic cues that presage such transitions by studying dyadic interactions in an online strategy game where players form alliances and break those alliances through betrayal. We characterize friendships that are unlikely to last and examine temporal patterns that foretell betrayal. We reveal that subtle signs of imminent betrayal are encoded in the conversational patterns of the dyad, even if the victim is not aware of the relationship's fate. In particular, we find that lasting friendships exhibit a form of balance that manifests itself through language. In contrast, sudden changes in the balance of certain conversational attributes---such as positive sentiment, politeness, or focus on future planning---signal impending betrayal.
Colloquium: Understanding Quantum Weak Values: Basics and Applications
Justin Dressel,Mehul Malik,Filippo M. Miatto,Andrew N. Jordan,Robert W. Boyd
Physics , 2013, DOI: 10.1103/RevModPhys.86.307
Abstract: Since its introduction 25 years ago, the quantum weak value has gradually transitioned from a theoretical curiosity to a practical laboratory tool. While its utility is apparent in the recent explosion of weak value experiments, its interpretation has historically been a subject of confusion. Here a pragmatic introduction to the weak value in terms of measurable quantities is presented, along with an explanation for how it can be determined in the laboratory. Further, its application to three distinct experimental techniques is reviewed. First, as a large interaction parameter it can amplify small signals above technical background noise. Second, as a measurable complex value it enables novel techniques for direct quantum state and geometric phase determination. Third, as a conditioned average of generalized observable eigenvalues it provides a measurable window into nonclassical features of quantum mechanics. In this selective review, a single experimental configuration to discuss and clarify each of these applications is used.
Streaming Variational Bayes
Tamara Broderick,Nicholas Boyd,Andre Wibisono,Ashia C. Wilson,Michael I. Jordan
Computer Science , 2013,
Abstract: We present SDA-Bayes, a framework for (S)treaming, (D)istributed, (A)synchronous computation of a Bayesian posterior. The framework makes streaming updates to the estimated posterior according to a user-specified approximation batch primitive. We demonstrate the usefulness of our framework, with variational Bayes (VB) as the primitive, by fitting the latent Dirichlet allocation model to two large-scale document collections. We demonstrate the advantages of our algorithm over stochastic variational inference (SVI) by comparing the two after a single pass through a known amount of data---a case where SVI may be applied---and in the streaming setting, where SVI does not apply.
Generalization of Equatorial Impact-Parameter Formulas for Rotating Bodies
James Graber
Physics , 2009,
Abstract: This paper computes co-rotating and contra-rotating impact-parameter formulas in the plane of symmetry for any plane symmetric and axisymmetric rotating body in all metric theories of gravity, including general relativity. Impact-parameter formulas are useful to compute the appearance of accreting black holes, neutron stars, and other emitting or reflecting matter near a gravitationally compact rotating body. These rotating-body impact-parameter formulas generalize similar impact-parameter formulas for the Kerr case derived by Bardeen and coworkers in 1972, and another general-metric formula for the spherical case published by Bodenner and Will in 2003.
A Robust Test of General Relativity in Space
James Graber
Physics , 2006, DOI: 10.1142/S0218271807011401
Abstract: LISA may make it possible to test the black-hole uniqueness theorems of general relativity, also called the no-hair theorems, by Ryan's method of detecting the quadrupole moment of a black hole using high-mass-ratio inspirals. This test can be performed more robustly by observing inspirals in earlier stages, where the simplifications used in making inspiral predictions by the perturbative and post-Newtonian methods are more nearly correct. Current concepts for future missions such as DECIGO and BBO would allow even more stringent tests by this same method. Recently discovered evidence supports the existence of intermediate-mass black holes (IMBHs). Inspirals of binary systems with one IMBH and one stellar-mass black hole would fall into the frequency band of proposed maximum sensitivity for DECIGO and BBO. This would enable us to perform the Ryan test more precisely and more robustly. We explain why tests based on observations earlier in the inspiral are more robust and provide preliminary estimates of possible optimal future observations.
Enumerative geometry of hyperelliptic plane curves
Tom Graber
Mathematics , 1998,
Abstract: We recursively compute the Gromov-Witten invariants of the Hilbert scheme of two points in the plane. By studying the space of stable maps and computing virtual contributions, we use these invariants to enumerate hyperelliptic plane curves of degree d and genus g passing through 3d+1 general points.
Page 1 /2440
Display every page Item

Copyright © 2008-2017 Open Access Library. All rights reserved.