Latent dirichlet allocation

doi:10.5555/944919.944937

Home
/
Papers
/
Latent dirichlet allocation

Journal Article•DOI•

Latent dirichlet allocation

David M. Blei¹, Andrew Y. Ng², Michael I. Jordan¹•Institutions (2)

University of California, Berkeley¹, Stanford University²

01 Mar 2003-Journal of Machine Learning Research (JMLR.org)-Vol. 3, pp 993-1022

TL;DR: This work proposes a generative model for text and other collections of discrete data that generalizes or improves on several previous models including naive Bayes/unigram, mixture of unigrams, and Hofmann's aspect model.

read less

Abstract: We describe latent Dirichlet allocation (LDA), a generative probabilistic model for collections of discrete data such as text corpora. LDA is a three-level hierarchical Bayesian model, in which each item of a collection is modeled as a finite mixture over an underlying set of topics. Each topic is, in turn, modeled as an infinite mixture over an underlying set of topic probabilities. In the context of text modeling, the topic probabilities provide an explicit representation of a document. We present efficient approximate inference techniques based on variational methods and an EM algorithm for empirical Bayes parameter estimation. We report results in document modeling, text classification, and collaborative filtering, comparing to a mixture of unigrams model and the probabilistic LSI model.

...read moreread less

Content maybe subject to copyright Report

Citations

PDF

Open Access

More filters

Proceedings Article•

Learning from Bullying Traces in Social Media

[...]

Jun-Ming Xu¹, Kwang-Sung Jun¹, Xiaojin Zhu¹, Amy Bellmore¹•Institutions (1)

University of Wisconsin-Madison¹

03 Jun 2012

TL;DR: Evidence is presented that social media, with appropriate natural language processing techniques, can be a valuable and abundant data source for the study of bullying in both worlds.

...read moreread less

Abstract: We introduce the social study of bullying to the NLP community. Bullying, in both physical and cyber worlds (the latter known as cyberbullying), has been recognized as a serious national health issue among adolescents. However, previous social studies of bullying are handicapped by data scarcity, while the few computational studies narrowly restrict themselves to cyberbullying which accounts for only a small fraction of all bullying episodes. Our main contribution is to present evidence that social media, with appropriate natural language processing techniques, can be a valuable and abundant data source for the study of bullying in both worlds. We identify several key problems in using such data sources and formulate them as NLP tasks, including text classification, role labeling, sentiment analysis, and topic modeling. Since this is an introductory paper, we present baseline results on these tasks using off-the-shelf NLP solutions, and encourage the NLP community to contribute better models in the future.

...read moreread less

329 citations

Proceedings Article•DOI•

Modeling User Exposure in Recommendation

[...]

Dawen Liang¹, Laurent Charlin², James McInerney¹, David M. Blei¹•Institutions (2)

Columbia University¹, McGill University²

11 Apr 2016

TL;DR: This paper proposes a new probabilistic approach that directly incorporates user exposure to items into collaborative filtering, and recovers one of the most successful state-of-the-art approaches as a special case of the model.

...read moreread less

Abstract: Collaborative filtering analyzes user preferences for items (e.g., books, movies, restaurants, academic papers) by exploiting the similarity patterns across users. In implicit feedback settings, all the items, including the ones that a user did not consume, are taken into consideration. But this assumption does not accord with the common sense understanding that users have a limited scope and awareness of items. For example, a user might not have heard of a certain paper, or might live too far away from a restaurant to experience it. In the language of causal analysis (Imbens & Rubin, 2015), the assignment mechanism (i.e., the items that a user is exposed to) is a latent variable that may change for various user/item combinations. In this paper, we propose a new probabilistic approach that directly incorporates user exposure to items into collaborative filtering. The exposure is modeled as a latent variable and the model infers its value from data. In doing so, we recover one of the most successful state-of-the-art approaches as a special case of our model (Hu et al. 2008), and provide a plug-in method for conditioning exposure on various forms of exposure covariates (e.g., topics in text, venue locations). We show that our scalable inference algorithm outperforms existing benchmarks in four different domains both with and without exposure covariates.

...read moreread less

329 citations

Cites background or methods from "Latent dirichlet allocation"

...LDA is shown using a white dot....
[...]
...Assuming there are K topics Φ = φ1:K , each of which is a categorical distribution over a fixed set of vocabulary, LDA treats each document as a mixture of these topics where the topic proportion xi is inferred from the data....
[...]
...We compare to collaborative topic regression (CTR), a state-of-the-art method for recommending scientific papers [30] combining both LDA and WMF.9 We did not compare with the more recent and scalable collaborative topic Poisson factorization (CTPF) [7] since the resulting performance differences may have been the result of CTPF Poisson likelihood (versus Gaussian likelihood for both ExpoMF and WMF)....
[...]
...One can understand LDA as representing documents in a low-dimensional “topic” space with the topic proportion xi being their coordinates....
[...]
..., word embeddings [17], latent Dirichlet allocation [2]), or the position of venue i obtained by first clustering all the venues in the data set then finding the expected assignment to L clusters for each venue....
[...]

Journal Article•DOI•

Cross-Modal Retrieval With CNN Visual Features: A New Baseline

[...]

Yunchao Wei¹, Yao Zhao¹, Canyi Lu², Shikui Wei¹, Luoqi Liu², Zhenfeng Zhu¹, Shuicheng Yan² - Show less +3 more•Institutions (2)

Beijing Jiaotong University¹, National University of Singapore²

01 Feb 2017-IEEE Transactions on Systems, Man, and Cybernetics

TL;DR: Off-the-shelf CNN visual features are extracted from the CNN model, which is pretrained on ImageNet with more than one million images from 1000 object categories, as a generic image representation to tackle cross-modal retrieval.

...read moreread less

Abstract: Recently, convolutional neural network (CNN) visual features have demonstrated their powerful ability as a universal representation for various recognition tasks. In this paper, cross-modal retrieval with CNN visual features is implemented with several classic methods. Specifically, off-the-shelf CNN visual features are extracted from the CNN model, which is pretrained on ImageNet with more than one million images from 1000 object categories, as a generic image representation to tackle cross-modal retrieval. To further enhance the representational ability of CNN visual features, based on the pretrained CNN model on ImageNet, a fine-tuning step is performed by using the open source Caffe CNN library for each target data set. Besides, we propose a deep semantic matching method to address the cross-modal retrieval problem with respect to samples which are annotated with one or multiple labels. Extensive experiments on five popular publicly available data sets well demonstrate the superiority of CNN visual features for cross-modal retrieval.

...read moreread less

329 citations

Cites background from "Latent dirichlet allocation"

...For text representation, we first obtain the feature vector based on 500 tokens (with stop words removed) and then the LDA model is used to compute the probability of each document under 100 topics....
[...]
...Specifically, many text feature extraction techniques, such as tf-idf and latent Dirichlet allocation (LDA) [2], can be employed to extract the input text features for TextNet....
[...]
...cation (LDA) [2], can be employed to extract the input text features for TextNet....
[...]
...For text features, we first extract the feature vector based on the 300 most frequent tokens (with stop words removed) and then utilize the LDA to compute the probability of each document under 100 topics....
[...]
...For text representation, we first obtain the feature vector based on 25 000 most frequent tokens (with stop words removed) and then use the LDA to compute the probability of each document under 1000 topics....
[...]

Proceedings Article•DOI•

Multilingual Models for Compositional Distributed Semantics

[...]

Karl Moritz Hermann¹, Phil Blunsom¹•Institutions (1)

University of Oxford¹

01 Jun 2014

TL;DR: The authors leverage parallel data and learn to strongly align the embeddings of semantically equivalent sentences, while maintaining sufficient distance between those of dissimilar sentences, without relying on word alignments or any syntactic information.

...read moreread less

Abstract: We present a novel technique for learning semantic representations, which extends the distributional hypothesis to multilingual data and joint-space embeddings. Our models leverage parallel data and learn to strongly align the embeddings of semantically equivalent sentences, while maintaining sufficient distance between those of dissimilar sentences. The models do not rely on word alignments or any syntactic information and are successfully applied to a number of diverse languages. We extend our approach to learn semantic representations at the document level, too. We evaluate these models on two cross-lingual document classification tasks, outperforming the prior state of the art. Through qualitative analysis and the study of pivoting effects we demonstrate that our representations are semantically plausible and can capture semantic relationships across languages without parallel data.

...read moreread less

329 citations

Journal Article•DOI•

Corporate funding and ideological polarization about climate change

[...]

Justin Farrell¹•Institutions (1)

Yale University¹

05 Jan 2016-Proceedings of the National Academy of Sciences of the United States of America

TL;DR: This research demonstrates how polarization efforts are influenced by a patterned network of political and financial actors, and highlights the important influence of private funding in public knowledge and politics, and provides researchers a methodological model for future studies that blend large-scale textual discourse with social networks.

...read moreread less

Abstract: Drawing on large-scale computational data and methods, this research demonstrates how polarization efforts are influenced by a patterned network of political and financial actors. These dynamics, which have been notoriously difficult to quantify, are illustrated here with a computational analysis of climate change politics in the United States. The comprehensive data include all individual and organizational actors in the climate change countermovement (164 organizations), as well as all written and verbal texts produced by this network between 1993–2013 (40,785 texts, more than 39 million words). Two main findings emerge. First, that organizations with corporate funding were more likely to have written and disseminated texts meant to polarize the climate change issue. Second, and more importantly, that corporate funding influences the actual thematic content of these polarization efforts, and the discursive prevalence of that thematic content over time. These findings provide new, and comprehensive, confirmation of dynamics long thought to be at the root of climate change politics and discourse. Beyond the specifics of climate change, this paper has important implications for understanding ideological polarization more generally, and the increasing role of private funding in determining why certain polarizing themes are created and amplified. Lastly, the paper suggests that future studies build on the novel approach taken here that integrates large-scale textual analysis with social networks.

...read moreread less

329 citations

Cites methods from "Latent dirichlet allocation"

...Analytical Approach To empirically examine these data in relationship to the research questions above regarding the influence of corporate funding on textual content, the study employs a combination of social network analysis and a form of large-scale computational text analysis called latent dirichlet allocation (LDA) (30) topic modeling....
[...]
...To empirically examine these data in relationship to the research questions above regarding the influence of corporate funding on textual content, the study employs a combination of social network analysis and a form of large-scale computational text analysis called latent dirichlet allocation (LDA) (30) topic modeling....
[...]
...Building on the standard LDA model (30), Roberts et al....
[...]
...Building on the standard LDA model (30), Roberts et al. (31) explain, importantly, that: As in LDA, each document arises as a mixture over K topics....
[...]
...Thus, there are three critical differences in the STM as compared to the LDA model. . .(1) topics can be correlated; (2) each document has its own prior distribution over topics, defined by covariate X rather than sharing a global mean; and (3) word use within a topic can vary by covariate U....
[...]

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
…
73
74
75
76
77
78
79
…
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200

Collapse

References

PDF

Open Access

More filters

Book•

Handbook of mathematical functions : with formulas, graphs, and mathematical tables

[...]

Milton Abramowitz, Irene A. Stegun

01 Jan 1970

17,608 citations

Book•

Bayesian Data Analysis

[...]

Andrew Gelman, John B. Carlin, Hal S. Stern, David B. Dunson, Aki Vehtari, Donald B. Rubin¹ - Show less +2 more•Institutions (1)

University of California, Irvine¹

01 Jan 1995

TL;DR: Detailed notes on Bayesian Computation Basics of Markov Chain Simulation, Regression Models, and Asymptotic Theorems are provided.

...read moreread less

Abstract: FUNDAMENTALS OF BAYESIAN INFERENCE Probability and Inference Single-Parameter Models Introduction to Multiparameter Models Asymptotics and Connections to Non-Bayesian Approaches Hierarchical Models FUNDAMENTALS OF BAYESIAN DATA ANALYSIS Model Checking Evaluating, Comparing, and Expanding Models Modeling Accounting for Data Collection Decision Analysis ADVANCED COMPUTATION Introduction to Bayesian Computation Basics of Markov Chain Simulation Computationally Efficient Markov Chain Simulation Modal and Distributional Approximations REGRESSION MODELS Introduction to Regression Models Hierarchical Linear Models Generalized Linear Models Models for Robust Inference Models for Missing Data NONLINEAR AND NONPARAMETRIC MODELS Parametric Nonlinear Models Basic Function Models Gaussian Process Models Finite Mixture Models Dirichlet Process Models APPENDICES A: Standard Probability Distributions B: Outline of Proofs of Asymptotic Theorems C: Computation in R and Stan Bibliographic Notes and Exercises appear at the end of each chapter.

...read moreread less

16,079 citations

"Latent dirichlet allocation" refers background in this paper

...Finally, Griffiths and Steyvers (2002) have presented a Markov chain Monte Carlo algorithm for LDA....
[...]
...Structures similar to that shown in Figure 1 are often studied in Bayesian statistical modeling, where they are referred to ashierarchical models(Gelman et al., 1995), or more precisely asconditionally independent hierarchical models(Kass and Steffey, 1989)....
[...]
...Structures similar to that shown in Figure 1 are often studied in Bayesian statistical modeling, where they are referred to as hierarchical models (Gelman et al., 1995), or more precisely as conditionally independent hierarchical models (Kass and Steffey, 1989)....
[...]

Journal Article•DOI•

Indexing by Latent Semantic Analysis

[...]

Scott Deerwester¹, Susan T. Dumais², George W. Furnas², Thomas K. Landauer², Richard A. Harshman³ - Show less +1 more•Institutions (3)

University of Chicago¹, Telcordia Technologies², University of Western Ontario³

01 Sep 1990-Journal of the Association for Information Science and Technology

TL;DR: A new method for automatic indexing and retrieval to take advantage of implicit higher-order structure in the association of terms with documents (“semantic structure”) in order to improve the detection of relevant documents on the basis of terms found in queries.

...read moreread less

Abstract: A new method for automatic indexing and retrieval is described. The approach is to take advantage of implicit higher-order structure in the association of terms with documents (“semantic structure”) in order to improve the detection of relevant documents on the basis of terms found in queries. The particular technique used is singular-value decomposition, in which a large term by document matrix is decomposed into a set of ca. 100 orthogonal factors from which the original matrix can be approximated by linear combination. Documents are represented by ca. 100 item vectors of factor weights. Queries are represented as pseudo-document vectors formed from weighted combinations of terms, and documents with supra-threshold cosine values are returned. initial tests find this completely automatic method for retrieval to be promising.

...read moreread less

12,443 citations

"Latent dirichlet allocation" refers methods in this paper

...To address these shortcomings, IR researchers have proposed several other dimensionality reduction techniques, most notably latent semantic indexing (LSI) (Deerwester et al., 1990)....
[...]
...To address these shortcomings, IR researchers have proposed several other dimensionality reduction techniques, most notablylatent semantic indexing (LSI)(Deerwester et al., 1990)....
[...]

Book•

Introduction to Modern Information Retrieval

[...]

Gerard Salton, Michael J. McGill

01 Jan 1983

TL;DR: Reading is a need and a hobby at once and this condition is the on that will make you feel that you must read.

...read moreread less

Abstract: Some people may be laughing when looking at you reading in your spare time. Some may be admired of you. And some may want be like you who have reading hobby. What about your own feel? Have you felt right? Reading is a need and a hobby at once. This condition is the on that will make you feel that you must read. If you know are looking for the book enPDFd introduction to modern information retrieval as the choice of reading, you can find here.

...read moreread less

12,059 citations

"Latent dirichlet allocation" refers background or methods in this paper

...In the populartf-idf scheme (Salton and McGill, 1983), a basic vocabulary of “words” or “terms” is chosen, and, for each document in the corpus, a count is formed of the number of occurrences of each word....
[...]
...We report results in document modeling, text classification, and collaborative filtering, comparing to a mixture of unigrams model and the probabilistic LSI model....
[...]

Book•

Theory of probability

[...]

Harold Jeffreys, R. Bruce Lindsay

01 Jan 1939

TL;DR: In this paper, the authors introduce the concept of direct probabilities, approximate methods and simplifications, and significant importance tests for various complications, including one new parameter, and various complications for frequency definitions and direct methods.

...read moreread less

Abstract: 1. Fundamental notions 2. Direct probabilities 3. Estimation problems 4. Approximate methods and simplifications 5. Significance tests: one new parameter 6. Significance tests: various complications 7. Frequency definitions and direct methods 8. General questions

...read moreread less

7,086 citations