Dynamic Topic Modeling – Meta-Guide.com

Notes:

Topic modeling is a set of techniques that are used to automatically discover the underlying topics in a dataset of text documents and to represent these topics in a concise and interpretable way. These techniques are typically used to understand the overall content and structure of a dataset, and they may not take into account the evolution of topics over time.

Dynamic topic modeling is a machine learning technique that is used to analyze and understand the evolution of topics over time in a dataset of text documents. It is a type of topic modeling, which is a set of techniques that are used to automatically discover the underlying topics in a dataset of text documents and to represent these topics in a concise and interpretable way.

Dynamic topic modeling is different from other types of topic modeling in that it is specifically designed to analyze the evolution of topics over time. This is accomplished by constructing a topic model for each time period in the dataset, and then using techniques such as visualization and clustering to analyze the relationships between topics across different time periods.

Other types of topic modeling, such as Latent Dirichlet Allocation (LDA), are typically used to discover the underlying topics in a dataset of text documents as a whole, without taking into account the evolution of topics over time. While these techniques can be useful for understanding the overall content and structure of a dataset, they may not be as effective at analyzing the dynamic changes in topics over time that are captured by dynamic topic modeling.

Dynamic topic mining is a process of discovering and analyzing the evolution of topics over time in a dataset of text documents. It typically involves constructing a topic model for each time period in the dataset, and then using techniques such as visualization and clustering to analyze the relationships between topics across different time periods. Dynamic topic mining is a type of dynamic topic modeling, which is a machine learning technique that is specifically designed to analyze the evolution of topics over time.
Topical phrase mining is a process of identifying and extracting phrases or terms that are related to a particular topic or theme from a dataset of text documents. These phrases or terms may be used to represent the topic or theme in a concise and interpretable way, and they may be used for tasks such as classification, visualization, or summarization. Topical phrase mining is often used in conjunction with topic modeling techniques to provide a more detailed and nuanced understanding of the content and structure of a dataset.

Resources:

github.com/Blei-Lab/turbotopics .. Turbo topics find significant multiword phrases in topics

Wikipedia:

References:

Mining Latent Entity Structures (2015)

Dependent hierarchical normalized random measures for dynamic topic modeling C Chen, N Ding, W Buntine – arXiv preprint arXiv:1206.4671, 2012 – arxiv.org Abstract: We develop dependent hierarchical normalized random measures and apply them to dynamic topic modeling. The dependency arises via superposition, subsampling and point transition on the underlying Poisson processes of these measures. The measures … Cited by 14 Related articles All 17 versions

Interactive topic modeling Y Hu, J Boyd-Graber, B Satinoff, A Smith – Machine learning, 2014 – Springer Page 1. Mach Learn (2014) 95:423–469 DOI 10.1007/s10994-013-5413-0 Interactive topic modeling Yuening Hu · Jordan Boyd-Graber · Brianna Satinoff · Alison Smith Received: 17 August 2012 / Accepted: 20 August 2013 … Cited by 81 Related articles All 33 versions

TM-LDA: efficient online modeling of latent topic transitions in social media Y Wang, E Agichtein, M Benzi – Proceedings of the 18th ACM SIGKDD …, 2012 – dl.acm.org Page 1. TM-LDA: Efficient Online Modeling of Latent Topic Transitions in Social Media Yu Wang Emory University yu.wang@emory.edu Eugene Agichtein Emory University eugene@mathcs. emory.edu Michele Benzi Emory University benzi@mathcs.emory.edu … Cited by 40 Related articles All 11 versions

Investigating drug repositioning opportunities in FDA drug labels through topic modeling H Bisgin, Z Liu, R Kelly, H Fang, X Xu… – BMC …, 2012 – biomedcentral.com … powerful drug repositioning system. Although the current study only considered the most recent drug labels, drug labels can also be mined at varying time points by using a dynamic topic modeling approach. In addition to predicting … Cited by 20 Related articles All 13 versions

Probabilistic topic models D Blei, L Carin, D Dunson – Signal Processing Magazine, IEEE, 2010 – ieeexplore.ieee.org Page 1. IEEE SIGNAL PROCESSING MAGAZINE [55] NOVEMBER 2010 1053-5888/10/$26.00© 2010IEEE IEEE SIGNAL PROCESSING MAGAZINE [55] NOVEMBER 2010 Digital Object Identifier 10.1109/MSP.2010.938079 [A focus on graphical model design … Cited by 48 Related articles All 11 versions

Learning evolving and emerging topics in social media: a dynamic nmf approach with temporal regularization A Saha, V Sindhwani – Proceedings of the fifth ACM international …, 2012 – dl.acm.org … parameters, LDA applies a Dirichlet prior on them. Variants of pLSI and LDA have been proposed for online and dynamic topic modeling (see [6, 13, 15, 4, 2] and references therein). Another line of seemingly unrelated work which … Cited by 53 Related articles All 6 versions

PatentMiner: topic-driven patent analysis and mining J Tang, B Wang, Y Yang, P Hu, Y Zhao, X Yan… – Proceedings of the 18th …, 2012 – dl.acm.org Page 1. PatentMiner: Topic-driven Patent Analysis and Mining Jie Tang†, Bo Wang†, Yang Yang†, Po Hu†, Yanting Zhao†, Xinyu Yan†, Bo Gao†, Minlie Huang†, Peng Xu?, Weichang Li?, and Adam K. Usadi? †Department … Cited by 28 Related articles All 6 versions

Emerging topic detection for organizations from microblogs Y Chen, H Amiri, Z Li, TS Chua – … of the 36th international ACM SIGIR …, 2013 – dl.acm.org … that have found remarkable success in building topic models of static text. Variants of PLSA and LDA have been pro- posed for online and dynamic topic modeling [4].Wang et al. [23] took advantage of temporal information, and tried to model the topics continuously over time. … Cited by 39 Related articles All 5 versions

Regularized latent semantic indexing: A new approach to large-scale topic modeling Q Wang, J Xu, H Li, N Craswell – ACM Transactions on Information …, 2013 – dl.acm.org … 31, No. 1, Article 5, Publication date: January 2013. Page 4. 5:4 Q. Wang et al. on dynamic topic modeling is empirically verified; and (3) a theoretical comparison of batch RLSI and online RLSI is given. The rest of the article is organized as follows. … Cited by 19 Related articles All 13 versions

Evolution in social networks: A survey M Spiliopoulou – Social Network Data Analytics, 2011 – Springer … As we see in the left box of the figure, community evolution is governed by Dimension 4. We discuss methods that observe a community as a crisp cluster in Subsection 4, and methods based on dynamic topic modeling and assuming temporal smoothness in Subsection 5. The … Cited by 32 Related articles All 6 versions

Real-time visualization of streaming text data: tasks and challenges C Rohrdantz, D Oelke, M Krstajic, F Fischer – 2011 – kops.uni-konstanz.de … world events. In the field of text mining, a lot of research has recently been done on online topic mining, with three main directions: Topic detection and tracking [2], dynamic topic modeling [5], and evolutionary clustering [25]. The … Cited by 17 Related articles All 8 versions

A Nonparametric Mixture Model for Topic Modeling over Time. A Dubey, A Hefny, S Williamson, EP Xing – SDM, 2013 – SIAM … (2012), is to sample the timestamps from a Dirichlet process mixture of Gaussians. However, this ignores the possibility of correlations between the trending patterns of topics, something that is not addressed in much of the dynamic topic modeling literature. … Cited by 8 Related articles All 8 versions

SolarMap: multifaceted visual analytics for topic exploration N Cao, D Gotz, J Sun, YR Lin… – Data Mining (ICDM), 2011 …, 2011 – ieeexplore.ieee.org … In this process, the position of node x of graph G depends on both the structure of G as well as the positions of x in G’s parent graphs. The layout method will preserve the stability of topic clusters over time, which can be used for visualizing any dynamic topic modeling results. IV. … Cited by 11 Related articles All 19 versions

A Dynamic Topic Model of Learning Analytics Research. M Derntl, N Günnemann, R Klamma – LAK (Data Challenge), 2013 – ceur-ws.org … a set of pa- pers. To identify what is relevant to LAK research, we used the dynamic topic modeling approach described in [3] to ob- tain the distribution of words over a pre-defined number of topics. This is a probabilistic, unsupervised … Cited by 2 Related articles

The dynamic features of Delicious, Flickr, and YouTube N Lin, D Li, Y Ding, B He, Z Qin, J Tang… – Journal of the …, 2012 – Wiley Online Library Skip to Main Content. Wiley Online Library. Log in / Register. Log In E-Mail Address Password Forgotten Password? Remember Me. … Cited by 16 Related articles All 8 versions

Discovering global and local bursts in a stream of news M Zimmermann, I Ntoutsi, ZF Siddiqui… – Proceedings of the 27th …, 2012 – dl.acm.org … Third, if the burst occurs inside a topic, ie it is a local burst, the keywords characteristic to it may never be frequent enough to become a topic of its own. Dynamic topic modeling constitutes a second family of ap- proaches on learning topics over a stream adaptively [12, 4, 15]. … Cited by 16 Related articles All 4 versions

An interactive system for visual analytics of dynamic topic models N Günnemann, M Derntl, R Klamma, M Jarke – Datenbank-Spektrum, 2013 – Springer … Abstract The vast amount and rapid growth of data on the Web and in document repositories make knowledge extrac- tion and trend analysis a challenging task. A well-proven approach for the unsupervised analysis of large text corpora is dynamic topic modeling. … Cited by 5 Related articles All 5 versions

D-VITA: A Visual Interactive Text Analysis System Using Dynamic Topic Mining. N Günnemann, M Jarke – BTW workshops, 2013 – btw-2013.de … database. 5 Conclusion In this paper, we present D-VITA, a novel interactive visual text analysis system based on dynamic topic modeling that is designed to support users exploring and interacting with numbers of documents. It … Cited by 2 Related articles All 5 versions

Visually Summarizing the Evolution of Documents under a Social Tag. A Gohr, M Spiliopoulou, A Hinneburg – KDIR, 2010 – informatik.uni-halle.de … We capture the evo- lution of topics by AdaptivePLSA (Gohr et al., 2009), an extension of PLSA for dynamic topic modeling. … To report top- ics learned by dynamic topic modeling, (Blei and Lafferty, 2006) list the most likely words for top- ics at several time points. … Cited by 5 Related articles All 4 versions

What can topic models of PMLA teach us about the history of literary scholarship A Goldstone, T Underwood – Journal of Digital Humanities, 2012 – tedunderwood.com … I have only experimented with dynamic topic modeling, which I think is interesting to compare with time-series data from regular LDA models (that’s what John Laudun, Clai Rice, and I are doing with folklore journals in a similar project we’re working on at the moment). … Cited by 5 Related articles All 2 versions

Theorizing research practices we forgot to theorize twenty years ago T Underwood – Representations, 2014 – JSTOR … m-rhody/. 15. Attempts to frame explicitly diachronic versions of topic modeling (like dynamic topic modeling and ”topics over time”) have tended to invoke dubi- ous assumptions about historical continuity. Historians are probably … Cited by 4 Related articles All 3 versions

Mining the Dispatch under Supervision: Using Casualty Counts to Guide Topics from the Richmond Daily Dispatch Cor C Templeton, T Brown, S Battacharyya… – Chicago Colloquium on …, 2011 – Citeseer … Mc- Callum, 2006). Dynamic Topic Modeling (Blei and Lafferty, 2006) allows topics to evolve from year to year, capturing the intuition that scientific fields, for example, endure despite changing terminology. Finally, Dirichlet Forests … Cited by 3 Related articles All 3 versions

Deconstruct and Reconstruct: Using Topic Modeling on an Analytics Corpus. M Sharkey, M Ansari – LAK Workshops, 2014 – ceur-ws.org … audience. 5.2 LAK13 Dynamic Topic Modeling Another 2013 LAK Data Challenge entrant, Derntl et. al. [6 … Topics. The researchers used Dynamic Topic Modeling, a precursor to the Turbo Topics technique developed by Blei. One … Cited by 1 Related articles

Dependent normalized random measures C Chen, V Rao, W Buntine, Y Whye Teh – Proceedings of The 30th …, 2013 – jmlr.org Page 1. Dependent Normalized Random Measures Changyou Chen1,3 Changyou.Chen@nicta.com.au Vinayak Rao2 vrao@gatsby.ucl.ac.uk Wray Buntine3,1 Wray.Buntine@nicta.com.au YeeWhye Teh4 ywteh@stats.ox.ac.uk … Cited by 6 Related articles All 11 versions

Sequential Summarization: A New Application for Timely Updated Twitter Trending Topics. D Gao, W Li, R Zhang – ACL (2), 2013 – Citeseer … It takes advantage of Dynamic Topic Modeling (David and Michael, 2006) to explore the tweet content. 568 Page 617. DTM in nature is a clustering approach which can dynamically generate the subtopic underlying the topic. … Cited by 5 Related articles All 8 versions

Sequential Summarization: A Full View of Twitter Trending Topics D Gao, W Li, X Cai, R Zhang… – Audio, Speech, and …, 2014 – ieeexplore.ieee.org … facing short-time topics. In order to locate each subtopic well, the semantic-based approach leverages Dynamic Topic Modeling … DTM, an extension of Latent Dirichlet Allocation (LDA) [20], is a generative model widely used in dynamic topic modeling researches. … Cited by 8 Related articles All 4 versions

Online egocentric models for citation networks H Wang, WJ Li – Proceedings of the Twenty-Third international joint …, 2013 – dl.acm.org Page 1. Online Egocentric Models for Citation Networks Hao Wang and Wu-Jun Li Shanghai Key Laboratory of Scalable Computing and Systems Department of Computer Science and Engineering, Shanghai Jiao Tong University … Cited by 3 Related articles All 6 versions

Hierarchical multi-label classification of social text streams Z Ren, MH Peetz, S Liang, W van Dolen… – Proceedings of the 37th …, 2014 – dl.acm.org … document x0ti ? X0ti Short text xti ? Xti … .. A) Document expansion (C) Chunk-based structural classification —- Entity linking with Wikipedia —- Query-based sentence ranking —- Dynamic topic modelling at ti —- Global topics z ? Zg ti —- Local topics z ? Zl ti … Cited by 7 Related articles All 5 versions

The emergence of the modern concept of introspection: a quantitative linguistic analysis I Raskovsky, DF Slezak, CG Diuk… – Proceedings of the NAACL …, 2010 – dl.acm.org … We will also incorporate the notion of concept drift to our topic modeling, expecting it to account for the temporal evolution of the use of introspec- tion. A promising proposal for this purpose is that of Dynamic Topic Modeling. … Cited by 2 Related articles All 17 versions

Evolution of Movie Topics Over Time C Meng, M Zhang, W Guo – 2012 – cs229.stanford.edu … Out of the 100 topics, around 10 topics were found more probable than others. Thus in Dynamic Topic Modeling (DTM), we picked 20 topics to fit our movie data. … 3 Page 4. Figure 3: Dynamic Topic Modeling (DTM) 20 topics model results. … Cited by 1 Related articles All 3 versions

Visually summarizing semantic evolution in document streams with topic table A Gohr, M Spiliopoulou, A Hinneburg – Knowledge Discovery, Knowledge …, 2013 – Springer … These topics may be learned by any dynamic topic modeling method; we use AdaptivePLSA [8], where a series of topic models is learned, and the topics learned at each timepoint are semantically linked to those appearing in the next timepoint. … Cited by 1 Related articles All 3 versions

Roles in social networks: Methodologies and research issues M Forestier, A Stavrianou, J Velcin… – Web Intelligence and …, 2012 – researchgate.net … The model pro- posed by the authors is based on the previous work on dynamic topic modeling [6] and static model that cap- ture role correlations [4]. More precisely, it augments the MMSB model with a state space model similar to that used in Dynamic Topic Models. … Cited by 22 Related articles All 7 versions

Modeling the Association Between Clusters of SNPs and Disease Responses R Argiento, A Guglielmi, CK Hsiao, F Ruggeri… – … Bayesian Inference in …, 2015 – Springer … See also Chen et al. (2012) for an application of such multivariate priors in a dynamic topic modeling context. How- ever, there is a recent and very lively literature on algorithms to draw inference for NGG-mixtures, which has resulted into a number of efficient algorithms. … Cited by 1

Incremental visual text analytics of news story development M Krstajic, M Najm-Araghi… – IS&T/SPIE …, 2012 – proceedings.spiedigitallibrary.org … Document clustering has been exhaustively researched in the field of text data mining, with several main directions, such as dynamic topic modeling based on LDA2 and topic detection and tracking.1 The evaluation of document clustering output can be quite exhausting and … Cited by 7 Related articles All 9 versions

Chronological Scientific Information Recommendation via Supervised Dynamic Topic Modeling Z Jiang – Proceedings of the Eighth ACM International …, 2015 – dl.acm.org Abstract Scientific information recommendation is crucial to assist scholars for their researches. Citation recommendation is an important field of scientific recommendation. Traditional approaches ignore the chronological nature of the citation recommendation … Related articles

Dynamic Topic Modeling for Monitoring Market Competition from Online Text and Image Data H Zhang, G Kim, EP Xing – Proceedings of the 21th ACM SIGKDD …, 2015 – dl.acm.org Abstract We propose a dynamic topic model for monitoring temporal evolution of market competition by jointly leveraging tweets and their associated images. For a market of interest (eg luxury goods), we aim at automatically detecting the latent topics (eg bags, clothes, …

A Bibliometric Analysis on Cancer Population Science with Topic Modeling DC Li, M Rastegar-Mojarad, J Okamoto… – AMIA Summits on …, 2015 – ncbi.nlm.nih.gov … We applied two topic modeling techniques: author topic modeling (AT) and dynamic topic modeling (DTM). Our initial results show that AT can address reasonably the issues related to investigators’ research interests, research topic distributions and popularities. …

Experiments with Dynamic Topic Models J Li, W Buntine – agava.ijs.si … period. Categories and Subject Descriptors I.7 [Document and Text Processing]: Miscellaneous; I.2.6 [Artificial Intelligence]: Learning Keywords dynamic topic modelling; experimental results; ABC News 1. INTRODUCTION … Related articles All 2 versions

Building and Exploring Dynamic Topic Models on the Web M Derntl, N Günnemann, A Tillmann… – Proceedings of the 23rd …, 2014 – dl.acm.org … documents. One of these techniques is called dynamic topic modeling [1], which uses algorithms that are able to identify the latent topics “hidden” in a set of documents and how these have evolved over time. While topic models can … Cited by 1 Related articles

Topic modelling in the information warfare domain A de Waal, F Mouton – Adaptive Science and Technology ( …, 2013 – ieeexplore.ieee.org … Using topic modelling, news articles can be grouped together based on their thematic similarity. Topic model variations such as dynamic topic modelling, author-topic models, supervised LDA can enrich the topic model output considerably. C. Facebook … Related articles All 4 versions

Bayesian Analysis of Dynamic Linear Topic Models C Glynn, ST Tokdar, DL Banks, B Howard – 2015 – stat.duke.edu Page 1. DRAFT Bayesian Analysis of Dynamic Linear Topic Models Chris Glynn1, Surya T. Tokdar1, David L. Banks1, and Brian Howard2 1Statistical Science, Duke University 2Sciome, LLC October 19, 2015 Abstract In dynamic …

Dynamic Bayesian activity modeling in video via multi-feature integration TS Brandes, E Wang – Acoustics, Speech and Signal …, 2013 – ieeexplore.ieee.org … 2. REVIEW OF DYNAMIC TOPIC MODELING … 3. FEATURES OBSERVED The type of distribution used to model observables and topics within the dynamic topic modeling framework is flexible, and should consist of whatever is most appropriate for the obser- vation itself. … Related articles

Telling New Stories about our Texts: Next Steps for Topic Modeling in the Humanities T Brown – Maryland Institute for Technology in the Humanities. …, 2012 – mith.umd.edu … Consumer price index as observation variable ? “Multilingual” SLDA model for North / South Page 45. Other extensions ? Dynamic topic modeling ? Topics over time ? Interactive topic modeling ? Syntactic topic models Page 46. Topic Modeling for Humanities Research … Related articles

Topic Models: A Tutorial with R GM Richardson, J Bowers, AJ Woodill… – … Journal of Semantic …, 2014 – World Scientific … structure in Twitter data. The twitter feeds in our applications do not lend to any temporal changes that suggests a document ordering, though see Sec. 5 for more discussion on dynamic topic modeling. [5] provides a detailed … Related articles

Citation Recommendation via Time-series Scholarly Topic Analysis and Publication Prior Analysis Z Jiang – ieee-tcdl.org … H.3.3 [Information Storage and Retrieval]: Information Search and Retrieval General Terms Algorithms, Measurement, Experimentation Keywords Bibliometrics, Citation Recommendation, Supervised Dynamic Topic Modeling, PageRank, Prior Knowledge 1. INTRODUCTION … Related articles

An expectation maximisation algorithm for behaviour analysis in video O Isupova, L Mihaylova, D Kuzin… – … (Fusion), 2015 18th …, 2015 – ieeexplore.ieee.org … In this paper we propose a novel approach for dynamic topic modeling to detect abnormal behaviour in video sequences. The topic model de- scribes activities and behaviours in the scene assuming behaviour temporal dynamics. …

Continuous-Time Infinite Dynamic Topic Models: The Dim Sum Process for Simultaneous W Elshamy, WH Hsu – Emerging Methods in Predictive Analytics: …, 2014 – books.google.com … Continuous-Time Infinite Dynamic Topic Models 1. DYNAMIC TOPIC MODELING AND EVENT STREAM MINING In this chapter, we discuss the problem of simul- taneous topic enumeration and tracking (STEF): that of maintaining both the number of topics and a parametric … All 2 versions

Visually summarizing the Evolution of Documents under a Social Tag (Resubmission) A Gohr, M Spiliopoulou, A Hinneburg – LWA 2010 – kde.cs.uni-kassel.de … To report top- ics learned by dynamic topic modeling, [Blei and Lafferty, 2006] list the most likely words for topics at several time points. Additionally, they plot the probability of certain words for a topic at different time points to give hints about how this topic changes through time. … Related articles All 2 versions

Deep Temporal Sigmoid Belief Networks for Sequence Modeling Z Gan, C Li, R Henao, D Carlson, L Carin – arXiv preprint arXiv: …, 2015 – arxiv.org … Two tasks are considered, ie, prediction and dynamic topic modeling. Prediction The prediction task is concerned with estimating the held-out words. We employ the setup in [31]. … Dynamic Topic Modeling The setup described in [30] is employed, and the number of topics is 200. …

Proceedings of the 29th International Conference on Machine Learning (ICML-12) J Langford, J Pineau – arXiv preprint arXiv:1207.4676, 2012 – arxiv.org … Balakrishnan, Min Xu, Aarti Singh. Pages: 887-894. Dependent Hierarchical Normalized Random Measures for Dynamic Topic Modeling Changyou Chen, Nan Ding, Wray Buntine. Pages: 895-902. State-Space Inference for … All 2 versions

Visualization of Clandestine Labs from Seizure Reports: Thematic Mapping and Data Mining Research Directions W Hsu, M Abduljabbar, R Osuga, M Lu… – arXiv preprint arXiv: …, 2015 – arxiv.org … topic modeling approaches. We develop a static, finite topic model and examine the potential benefits and feasibility of extending this to dynamic topic modeling with a large number of topics and continuous tome. We describe … Related articles All 7 versions

Intelligent Topical Sentiment Analysis for the Classification of E-Learners and Their Topics of Interest M Ravichandran, G Kulanthaivel… – The Scientific World …, 2015 – hindawi.com The Scientific World Journal is a peer-reviewed, open access journal covering a wide range of subjects in science, technology, and medicine. The journal’s Editorial Board as well as its Table of Contents are divided into 98 subject areas that are covered within the journal’s scope … Related articles All 7 versions

Chronological Citation Recommendation with Information-Need Shifting Z Jiang, X Liu, L Gao – Proceedings of the 24th ACM International on …, 2015 – dl.acm.org

Complex Politics: A Quantitative Semantic and Topological Analysis of UK House of Commons Debates S Gurciullo, M Smallegan, M Pereda, F Battiston… – arXiv preprint arXiv: …, 2015 – arxiv.org … By the use of dynamic topic modeling (DTM) and topological data analysis (TDA) we show that both members and parties feature specific roles within the system, consistent over time, and extract global patterns indicating levels of political cohesion. …

Modeling the Flow and Change of Information on the Web N Pobiedina – Proceedings of the 21st international conference …, 2012 – dl.acm.org … evolutionary modeling of social networks; dynamic topic modeling; and graph transformation systems. The … quality. In the area of dynamic topic modeling scientists solve problems of topic identification and tracking over time. In … Related articles All 7 versions

Structure and Trends in a Selection of Academic Literature H Hironaka, R Suri – Citeseer … where the numerator is the joint PDF and the denominator is the probability of the evidence. The two main approaches for numerically approximating the posterior are Gibbs sampling [7, 12] and variational methods [10, 14]. Dynamic Topic Modeling and Extensions to LDA. … Related articles All 2 versions

Discovery of Rare Sequential Topic Patterns in Document Stream Z Hu, H Wang, J Zhu, M Li, Y Qiao, C Deng – SIAM … as a text stream. To obtain the temporal dy- namics of topics, various dynamic topic modeling meth- ods have been proposed to discover topics over time in document streams [6, 18, 14, 24, 27]. However, these methods were … Related articles

Accounting for Language Changes over Time in Document Similarity Search S Morsy, G Karypis – 2015 – cs.umn.edu … query given in today’s language in order to translate it to the user-specified date’s language so that the translated query is then used to retrieve old docu- ments relevant to the query [16, 8, 9, 10, 7]. Another line of research developed dynamic topic modeling approaches to …

Establishing Video Game Genres Using Data-Driven Modeling and Product Databases A Faisal, M Peltoniemi – Games and Culture, 2015 – gac.sagepub.com … documents. Extensions of multitask LDA models toward dynamic topic modeling offers opportunities for examining changes in the social construction of genres over time and may reveal interesting innovation trends in game content. …

A Temporal Expert Finding Methodology Based On United Author-Document-Topic Graphs AE KILINÇ – 2014 – etd.lib.metu.edu.tr … AVGP Average Precision DTM Dynamic Topic Modeling GLOSSEX Glossary Extraction HITS Hyperlink-Induced Topic Search … eling techniques. Besides, some of these studies use dynamic topic modeling tech- niques and extensions to Dynamic LDA. … Related articles

ISWC 2013 Doctoral Consortium ‘Ontology Evolution for End-User Communities’ PJ Goodall, P Eklund – Citeseer … metadata. The time-trace of these tags bears an interesting affinity with Dynamic Topic Modeling and Topic Hierarchies [2,3,12] – an approach for incremental discovery of hierarchic categories which appears worth investigating. … Related articles All 3 versions

A Latent Space Analysis of Editor Lifecycles in Wikipedia X Qin, D Greene, P Cunningham – arXiv preprint arXiv:1407.7736, 2014 – arxiv.org … This paper complements the previous works by characterizing the evolution of user activity in online communities using dynamic topic modeling, examines the patterns of change in users as modeled by their activity in the latent repre- Page 4. 4 X. Qin, D. Greene, P. Cunningham … Related articles All 4 versions

Emergence and Evolution of Topics in Social Media A Saha, V Sindhwani – evemnmf.googlecode.com … parameters, LDA applies a Dirichlet prior on them. Variants of pLSI and LDA have been proposed for online and dynamic topic modeling (see [6, 11, 13, 4, 2] and references therein). Another line of seemingly unrelated work which … Related articles All 2 versions

Unveiling the Political Agenda of the European Parliament Plenary: A Topical Analysis D Greene, JP Cross – arXiv preprint arXiv:1505.07302, 2015 – arxiv.org … 3.2 Dynamic Topic Modeling When applying clustering to temporal data, authors have often pro- posed dividing the data into time windows of fixed duration [24]. … We subsequently applied dynamic topic modeling as de- scribed in Section 3.2. … Cited by 1 Related articles All 2 versions

A revised inference for correlated topic model T Masada, A Takasu – Advances in Neural Networks–ISNN 2013, 2013 – Springer … JMLR 3, 993–1022 (2003) 3. Blei, D., Lafferty, J.: Correlated topic models. NIPS 18, 147–154 (2006) 4. Chen, C., Ding, N., Buntine, W.: Dependent hierarchical normalized random mea- sures for dynamic topic modeling. In: Proc. … Related articles All 4 versions

Emerging Methods in Predictive Analytics: Risk Management and Decision WH Hsu – 2014 – people.cis.ksu.edu … William H. Hsu Kansas State University, USA Page 4. 188 Continuous-Time Infinite Dynamic Topic Models 1. DYNAMIC TOPIC MODELING AND EVENT STREAM MINING … Both evolve in continuous time. 1.1 Problem Overview: Goals of Static and Dynamic Topic Modeling … All 2 versions

Human Rights Event Detection from Heterogeneous Social Media Graphs F Chen, DB Neill – Big Data, 2015 – online.liebertpub.com … Our future work will provide additional support for sensemaking and storytelling by extending NPHGS with approaches such as dynamic topic modeling, 10,11 which can provide more detailed analysis of tweet content and draw connections between related clusters over time. … Related articles All 5 versions

Ontology-based top-k query answering over massive, heterogeneous, and dynamic data? D Dell’Aglio – Doctoral (Consortium, 2013 – Citeseer … metadata. The time-trace of these tags bears an interesting affinity with Dynamic Topic Modeling and Topic Hierarchies [2, 3, 12]-an approach for incremental discovery of hierarchic categories which appears worth investigating. … Related articles All 4 versions

Computing Folklore Studies: Mapping over a Century of Scholarly Production through Topics J Laudun, J Goodwin – Journal of American Folklore, 2013 – muse.jhu.edu … that make up topics over time. Blei and JD Lafferty developed a variant algorithm known as “dynamic topic modeling” that seeks to account for changes in topics over time (Blei and Lafferty 2006). Changes in the distribution of … Related articles All 7 versions

On Generative Models For Sequential Formation Of Clusters PM Djuric, K Yu – eurasip.org … The processes that we study in this paper belong to this class. An early work on this idea in the context of dynamic topic modeling is [8]. In [9], the authors study evolutionary clustering over epochs with the aim of making the clustering parameters smooth over time. …

A survey of non-exchangeable priors for Bayesian nonparametric models NJ Foti, S Williamson – arXiv preprint arXiv:1211.4798, 2012 – arxiv.org Page 1. A survey of non-exchangeable priors for Bayesian nonparametric models Nicholas J. Foti1 and Sinead Williamson2 1Department of Computer Science, Dartmouth College 1Machine Learning Department, Carnegie Mellon University November 21, 2012 … Cited by 1 Related articles All 2 versions

Topic Uncovering and Image Annotation via Scalable Probit Normal Correlated Topic Models X Yu – 2015 – scholarworks.rit.edu … different time span. In the case of modeling changing topic overtime and predicting future topics also known as the dynamic topic modeling [13], the ordering of the document has to be taken into account as now a topic is an ordered sequence of words. A … Related articles

Time series modeling with hidden variables and gradient-based algorithms P Mirowski – 2011 – Citeseer … sum- mer support for my work on sentiment analysis and text categorization, and who sparked my research interest in dynamic topic modeling. And last but not least among the external collaborations, I am very grateful to Sumit … Cited by 5 Related articles All 10 versions

Patterns in the Emergence of Nanotechnology: The Case of Fullerenes S Kaplan, K Vakili – mackinstitute.wharton.upenn.edu Page 1. Patterns in the Emergence of Nanotechnology: The Case of Fullerenes Sarah Kaplan* University of Toronto, Rotman School 105 St. George Street Toronto, ON, M5S 3E6, Canada 416-978-7403 sarah.kaplan@rotman.utoronto.ca … Related articles

Customer relationship management and Web mining: the next frontier A Tuzhilin – Data Mining and Knowledge Discovery, 2012 – Springer Page 1. Data Min Knowl Disc (2012) 24:584–612 DOI 10.1007/s10618-012-0256- z Customer relationship management and Web mining: the next frontier Alexander Tuzhilin Received: 25 May 2010 / Accepted: 20 January 2012 … Cited by 18 Related articles All 8 versions

Twitter in academic events: A study of temporal usage, communication, sentimental and topical patterns in 16 Computer Science conferences D Parra, C Trattner, D Gómez, M Hurtado, X Wen… – Computer …, 2015 – Elsevier … In this context, a popular approach is called the Dynamic Topic Modeling (DTM) [5], this particular model assumes that the order of the documents reflects an evolving set of topics. Hence, it identifies a fixed set of topics that evolve over a period of time. … Cited by 1

Time Aware Knowledge Extraction for Microblog Summarization on Twitter C De Maio, G Fenza, V Loia, M Parente – arXiv preprint arXiv:1501.06715, 2015 – arxiv.org … peak areas according to the timestamps of the tweets; and a semantic based approach leveraging on Dynamic Topic Modeling, that extends LDA in order to consider timeline, to identify topic from a semantic prospective in the time interval. … Related articles All 2 versions

Dynamic Infinite Mixed-Membership Stochastic Blockmodel. X Fan, L Cao, RY Xu, R Yi – arXiv preprint arXiv:1306.2999, 2013 – ieeexplore.ieee.org Page 1. This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 1 Dynamic Infinite Mixed-Membership … Related articles All 5 versions

That’s sick dude!: Automatic identification of word sense change across different timescales S Mitra, R Mitra, M Riedl, C Biemann… – arXiv preprint arXiv: …, 2014 – arxiv.org … to events that are characterized mostly by named entities. Other similar works on dynamic topic modelling can be found in (Blei and Laf- ferty, 2006; Wang and McCallum, 2006). Google books n-gram viewer1 is a phrase-usage … Cited by 5 Related articles All 13 versions

An automatic approach to identify word sense changes in text media across timescales S MITRA, R MITRA, SK MAITY, M RIEDL… – Natural Language … – Cambridge Univ Press … Blei and Lafferty (2006) used Page 4. 4 S. Mitra et al. a dataset spanning 100 years from Science and using dynamic topic modeling, to analyze the time evolution of topics. … In dynamic topic modeling, the distribution of words associated with a topic change over time. … Related articles All 2 versions

Lexical shifts, substantive changes, and continuity in State of the Union discourse, 1790–2014 A Rule, JP Cointet, PS Bearman – Proceedings of the …, 2015 – National Acad Sciences … exogenous events. However, ToT is not designed to capture how a changing set of words or terms compose a topic over time. Blei and Lafferty’s (26) continuous dynamic topic modeling (cDTM) does do this. Like LDA, cDTM …

Knowledge Extraction and Reuse within” Smart” Service Centers C Wang, R Akella, S Ramachandran… – … (SRII), 2011 Annual, 2011 – ieeexplore.ieee.org Page 1. Knowledge Extraction and Reuse within “Smart” Service Centers Chunye Wang, Ram Akella Technology and Information Management University of California Santa Cruz Santa Cruz, USA {cwang, akella}@soe.ucsc.edu … Cited by 2 Related articles All 5 versions

Incremental Knowledge Discovery in Social Media X Tang – 2013 – idea.library.drexel.edu … model its topic-word distribution while propagating the hyper-parameter of topic distribution to the next incoming document. DMM is different from all previous dynamic topic modeling techniques in two main aspects. On one hand, it avoids the risk of drawing all timestamps … Related articles All 4 versions

Discovering and monitoring product features and the opinions on them with OPINSTREAM M Zimmermann, E Ntoutsi, M Spiliopoulou – Neurocomputing, 2015 – Elsevier … dimensions, adding new words and forgetting obsolete ones. Among dynamic topic modeling methods, there are also few that allow for changes in the set of dimensions [17] and [18]. Our earlier text stream clustering algorithm TStream … Cited by 3 Related articles All 2 versions

A survey of non-exchangeable priors for Bayesian nonparametric models NJ Foti, S Williamson – Pattern Analysis and Machine …, 2015 – ieeexplore.ieee.org Page 1. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL. 37, NO. 2, FEBRUARY 2015 359 A Survey of Non-Exchangeable Priors for Bayesian Nonparametric Models Nicholas J. Foti and Sinead A. Williamson … Cited by 3 Related articles All 5 versions

Accelerated Learning Of Latent Tree Models For Topic Detection And Multidimensional Clustering LIU TENGFEI – 2015 – cse.ust.hk Page 1. ACCELERATED LEARNING OF LATENT TREE MODELS FOR TOPIC DETECTION AND MULTIDIMENSIONAL CLUSTERING by LIU TENGFEI A Thesis Submitted to The Hong Kong University of Science and Technology … Related articles

Scientific Information Understanding via Open Educational Resources (OER) and Information Need Characterization X Liu, Z Jiang, L Gao – researchgate.net … For instance, as Figure 1 shows, user highlighted text is focusing on “unsupervised topic modeling for citation analysis”, and the paper content mainly addresses “dynamic topic modeling and PageRank-based citation recommendation”. …

An Automatic Similarity Detection Engine Between Sacred Texts Using Text Mining and Similarity Measures SHM Qahl – 2014 – scholarworks.rit.edu Page 1. ROCHESTER INSTITUTE OF TECHNOLOGY An Automatic Similarity Detection Engine Between Sacred Texts Using Text Mining and Similarity Measures by Salha Hassan Muhammed Qahl Supervisor: Professor Ernest Fokoue … Related articles All 2 versions

Theory of dependent hierarchical normalized random measures C Chen, W Buntine, N Ding – arXiv preprint arXiv:1205.4159, 2012 – arxiv.org Page 1. arXiv:1205.4159v2 [cs.LG] 25 May 2012 Theory of Dependent Hierarchical Normalized Random Measures Changyou Chen1,2 cchangyou@gmail.com 1RSISE, The Australian National University, Canberra, ACT, Australia Wray Buntine2,1 … Cited by 3 Related articles All 5 versions

Identification and on-line incremental clustering of spam campaigns C Sarantopoulos – 2015 – dspace.library.uu.nl Page 1. Identification and on-line incremental clustering of spam campaigns Charalampos Sarantopoulos August 2015 Master thesis Supervisors: Dr. AJ Feelders Drs. MJ den Uyl V. Hoekstra Page 2. Acknowledgments For this …

Word prediction techniques for user adaptation and sparse data mitigation K Trnka – 2010 – cis.udel.edu Page 1. WORD PREDICTION TECHNIQUES FOR USER ADAPTATION AND SPARSE DATA MITIGATION by Keith Trnka A dissertation submitted to the Faculty of the University of Delaware in partial fulfillment of the requirements … Cited by 2 Related articles All 8 versions

Mapping Microblogs into A Network of Topics: A case study on Microblogs generated in learning Activities G Abu-Oda – library.iugaza.edu.ps Page 1. Page 2. Islamic University of Gaza Faculty of Information Technology “Mapping Microblogs into A Network of Topics: A case study on Microblogs generated in learning Activities” By Ghadeer Abu-Oda Master Thesis A Master Thesis presented to the Faculty of Information …

A Bayesian nonparametric model for density and cluster estimation: the ?-NGG process mixture model R Argiento, I Bianchini, A Guglielmi – researchgate.net … distributions. See also Chen et al. (2012) for an application of such multivariate priors in a dynamic topic modeling context. 3 ?-NGG processes The goal of this section is the definition of a finite dimensional random probability measure … Related articles

Visual Analytics of Temporal Event Sequences in News Streams M Krstajic – 2014 – kops.uni-konstanz.de Page 1. Visual Analytics of Temporal Event Sequences in News Streams Dissertation zur Erlangung des akademischen Grades des Doktors der Naturwissenschaften an der Universität Konstanz im Fachbereich Informatik und Informationswissenschaft vorgelegt von … Related articles All 4 versions

Incremental tree-based inference with dependent normalized random measures J Lee, S Choi – Proceedings of the International Conference on …, 2014 – jmlr.csail.mit.edu Page 1. Incremental Tree-Based Inference with Dependent Normalized Random Measures Juho Lee and Seungjin Choi Department of Computer Science and Engineering Pohang University of Science and Technology 77 … Cited by 1 Related articles All 7 versions

Patent Mining: A Survey L Zhang, L Li, T Li – ACM SIGKDD Explorations Newsletter, 2015 – dl.acm.org Page 1. Patent Mining: A Survey Longhui Zhang lzhan015@cs.fiu.edu Lei Li lli003@cs.fiu.edu Tao Li taoli@cs.fiu.edu School of Computing and Information Sciences Florida International University Miami, FL 33199 ABSTRACT … Cited by 1 Related articles All 3 versions

A blocked Gibbs sampler for NGG-mixture models via a priori truncation R Argiento, I Bianchini, A Guglielmi – Statistics and Computing, 2015 – Springer … Griffin et al. (2013) and Lijoi et al. (2014) propose a vector of dependent NGG processes for comparing distribu- tions. See also Chen et al. (2012) for an application of such multivariate priors in a dynamic topic modeling context. 3 ?-NGG processes … Cited by 3 Related articles

Supplementary material for dependent normalized random measures C Chen, V Rao, W Buntine, YW Teh – Proceedings of the International …, 2013 – jmlr.org Page 1. Supplementary material for Dependent Normalized Random Measures Changyou Chen1,3 Changyou.Chen@nicta.com.au 1RSISE, Australian National University, Australia Vinayak Rao2 vrao@gatsby.ucl.ac.uk 2Dept. Statistical Science, Duke University, USA … Cited by 1 Related articles All 8 versions