Probabilistic Graphical Models & Dialog Systems


Probabilistic Graphical Models & Dialog Systems


Using automatically transcribed dialogs to learn user models in a spoken dialog system [PDF] from aclweb.org U Syed… – Proceedings of the 46th Annual Meeting of the …, 2008 – dl.acm.org … es- timate: for example, ASR models that rely a simple confusion rate and uniform substitutions (which can be estimated from small number of transcriptions) have been used to train dialog systems which out … Figure 1: A probabilistic graphical model of a human- computer dialog. … Cited by 10 – Related articles – All 18 versions

[PDF] Toward interpreting spatial language discourse with grounding graphs [PDF] from mit.edu D Simeonov, S Tellex, T Kollar… – 2011 RSS Workshop on …, 2011 – projects.csail.mit.edu … In order to tackle it we first construct a probabilistic graphical model for each sentence which encodes the meaning as a probability distribution. The topology of the Page 2. … Partially observable markov decision processes for spoken dialog systems. Comput. … Cited by 1 – Related articles – View as HTML – All 2 versions

Probabilistic ontology trees for belief tracking in dialog systems [PDF] from oregonstate.edu N Mehta, R Gupta, A Raux… – Proceedings of the 11th …, 2010 – dl.acm.org … Probabilistic Ontology Trees for Belief Tracking in Dialog Systems Neville Mehta Oregon State University mehtane@eecs.oregonstate.edu … Abstract We introduce a novel approach for robust belief tracking of user intention within a spoken dialog system. … Cited by 2 – Related articles – All 12 versions

Hybrid approach to user intention modeling for dialog simulation [PDF] from aclweb.org S Jung, C Lee, K Kim… – Proceedings of the ACL-IJCNLP 2009 …, 2009 – dl.acm.org … The noisy utterance is passed to a dialog system which consists of spoken language understanding (SLU) and dialog management (DM) modules. … A Markov Logic Network (MLN) combines first-order logic and probabilistic graphical models in a single representation. … Cited by 6 – Related articles – All 16 versions

Grammatical error simulation for computer-assisted language learning S Lee, J Lee, H Noh, K Lee… – Knowledge-Based Systems, 2011 – Elsevier … Recently, user simulation has become widely used in the development of spoken dialog systems to develop dialog strategies that use reinforcement learning (Scheffler & … The MLN is a template for the construction of probabilistic graphical models, namely Markov random fields. … Cited by 2 – Related articles – All 3 versions

Goal Recognition with Markov Logic Networks for Player-Adaptive Games [PDF] from ncsu.edu EY Ha, JP Rowe, BW Mott… – Seventh Artificial Intelligence and …, 2011 – aaai.org … plan recognition is used to infer characters’ goals from their actions (Charniak and Goldman 1993); in dialogue systems, it supports … MLNs encode undirected probabilistic graphical models with structures that are determined by first-order logic formulae and associated weights. … Related articles – All 3 versions

Unified treatment of data-sparseness and data-overfitting in maximum entropy modeling F Weng… – EP Patent 1,783,744, 2011 – freepatentsonline.com … As more applications are included spoken dialog systems, the limited usage of spoken language imposed by the systems, shallow understanding, and/or … 2, No 3, p. 290 – 294, 2003 shows a probabilistic graphical model for language processing using Gaussian priors. … Related articles – Cached

Hybrid user intention modeling to diversify dialog simulations [PDF] from postech.ac.kr S Jung, C Lee, K Kim, D Lee… – Computer Speech & Language, 2011 – Elsevier … When the dialog logs contain enough dialogs, the collected dialogs are used to evaluate the performance of the simulator and dialog systems, and to learn … A Markov Logic Network (MLN) combines probabilistic graphical models and first-order logic in a single representation. … Cited by 1 – Related articles – All 6 versions

[PDF] Belief modelling for situation awareness in human-robot interaction [PDF] from uio.no P Lison, C Ehrler… – … of the 19th International Symposium on …, 2010 – folk.uio.no … ism. Markov Logic is a combination of first-order logic and probabilistic graphical models. … 2 Markov Logic Networks Markov logic combines first-order logic and probabilistic graphical models in a unified representation [24]. From … Cited by 7 – Related articles – View as HTML – All 11 versions

SYSTEM AND METHOD FOR GENERATING USER MODELS FROM TRANSCRIBED DIALOGS J Williams, U Syed – US Patent App. 12/552,832, 2009 – Google Patents … 2 illustrates a functional block diagram of an exemplary natural language spoken dialog system; [0017] FIG. 3 illustrates an exemplary method embodi- ment; and [0018] FIG. 4 illustrates a probabilistic graphical model of a human-computer dialog. … Cited by 1 – All 2 versions

Complementary computing: policies for transferring callers from dialog systems to human receptionists [PDF] from microsoft.com E Horvitz… – User Modeling and User-Adapted Interaction, 2007 – Springer … Complementary computing: policies for transferring callers from dialog systems to human receptionists Eric Horvitz · Tim Paek … We focus specifically on the challenge of determining when it is best to transfer callers from an automated dialog system to human receptionists. … Cited by 19 – Related articles – BL Direct – All 8 versions

Bayesian network-based behavior control for skilligent robots [PDF] from hanyang.ac.kr SH Lee… – Robotics and Automation, 2009. ICRA’09. …, 2009 – ieeexplore.ieee.org … They focused on modeling error-handling pro- cesses in spoken dialogue systems, which include complex combinations of different multimodal … A. Parameter Learning of Hybrid Bayesian Network We use a probabilistic graphical model known as a Bayesian network to encode … Cited by 9 – Related articles – All 3 versions

Combined intention, activity, and motion recognition for a humanoid household robot [PDF] from kit.edu D Gehrig, P Krauthausen, L Rybok… – … and Systems (IROS …, 2011 – ieeexplore.ieee.org … plan recognition, have been successfully employed, eg, in software agents [5]. Probabilistic intention recognition with probabilistic graphical models has been developed … The system will be integrated with a larger multi-modal dialog system and will become part of the humanoid … Cited by 1 – Related articles – All 7 versions

[PDF] Transformation-based Learning for Semantic parsing [PDF] from pitt.edu F Jurcicek, M Gašic, S Keizer, F Mairesse… – 2009 – cs.pitt.edu … Note that modern statisti- cal dialogue systems typically exploit multiple ASR hypothe- ses. … Third, Markov Logic Networks (MLN) have been used to extract slot values by combining probabilistic graphical models and first-order logic [11]. … Cited by 2 – Related articles – All 14 versions

3rd International Conference on Affective Computing and Intelligent Interaction-ACII 2009 [PDF] from utwente.nl J Cohn, A Nijholt… – 2009 – eprints.eemcs.utwente.nl … Prof. Ji’s research interests are in computer vision, pattern recognition, and probabilistic graphical models. … 678 Emotion Detection in Dialog Systems: Applications, Strategies and Challenges Felix Burkhardt, Markus van Ballegooy, Klaus-Peter Engelbrecht, Tim Polzehl, Joachim … All 6 versions

Dynamic Bayesian networks and variable length genetic algorithm for designing cue-based model for dialogue act recognition AA Yahya, R Mahmod… – Computer Speech & Language, 2010 – Elsevier … tickets. Obviously, such dialogue system is inadequate. … extensions. 2. DBNs for DAR. Bayesian network (BN) is a probabilistic graphical model for representing and reasoning on uncertain domains (Korb and Nicholson, 2004). … Related articles – All 5 versions

Towards relational POMDPs for adaptive dialogue management [PDF] from aclweb.org P Lison – Proceedings of the ACL 2010 Student Research …, 2010 – dl.acm.org … Furthemore, the dialogue system also needs to be adaptive to its user (at- tributed beliefs and intentions, attitude, attentional state) and to … Markov Logic combines first-order logic and probabilistic graphical models in a unified repre- sentation (Richardson and Domingos, 2006). … Cited by 1 – Related articles – All 21 versions

[PDF] Fusing symbolic and decision-theoretic problem solving+ perception in a graphical cognitive architecture [PDF] from usc.edu J CHEN, A DEMSKI, T HAN… – Proceedings of the …, 2011 – pollux.usc.edu … a neologism for conditions and actions – meld these functionalities, by passing messages both from and to WM, to yield the bidirectional processing that is crucial in probabilistic graphical models. … In Proceedings of the AI-ED 2001 Workshop on Tutorial Dialogue Systems, 2001. … Cited by 2 – Related articles – View as HTML – All 3 versions

Contrasting the interaction structure of an email and a telephone corpus: A machine learning approach to annotation of dialogue function units [PDF] from psu.edu J Hu, RJ Passonneau… – … : The 10th Annual Meeting of the …, 2009 – dl.acm.org … Recently, discrimi- native Probabilistic Graphical Models have been widely applied in structural problems (Getoor and 360 Page 5. Taskar, 2007) such as link prediction. … 2008. Using dialogue acts to learn better repair strategies for spoken dialogue systems. In ICASSP. … Cited by 4 – Related articles – All 15 versions

Basic tutorial tactics for virtual agents [PDF] from uva.nl M Wißner, E André, M Häring, G Mehlmann… – 2010 – dare.uva.nl … DynaLearn D5.2 Abstract This document presents the progress and effort made in the design of a dialog system for the virtual characters in DynaLearn. … AutoTutor follows a 5-step dialog system, a pattern Graesser et al. retrieved from their observations of human expert tutors. … Cited by 2 – Related articles – All 2 versions

Probabilistic Data Integration [PDF] from utwente.nl M van Keulen – 2009 – eprints.eemcs.utwente.nl … We have developed a uniform, closed framework for representing and querying uncertain data based on concepts from probabilistic graphical models; I will present an overview of our framework, the … Contextual Information is essential for adaptive intelligent dialogue systems. … All 5 versions

[PDF] Bayesian Recommender Systems: Models and Algorithms [PDF] from anu.edu.au S Guo – 2011 – users.cecs.anu.edu.au Page 1. Bayesian Recommender Systems: Models and Algorithms Shengbo Guo October 2011 A thesis submitted for the degree of Doctor of Philosophy of The Australian National University Page 2. Page 3. Declaration These … Related articles – View as HTML

[PDF] Parsing PCFG within a general probabilistic inference framework [PDF] from agi-conf.org A Murugesan… – Proc. AGI, 2009 – agi-conf.org … Exact Inference Over GenProb Many of the existing approaches for combining first- order logic and probabilistic graphical models proposi- tionalize relational theories and making inferences over these … Mixed-Initiative Dialogue Systems for Collaborative Problem-Solving. … Cited by 2 – Related articles – View as HTML – All 7 versions

Transformation-Based Learning for Semantic Parsing F Jurcícek, M Gašic, S Keizer… – … Annual Conference of …, 2009 – isca-speech.org … Note that modern statisti- cal dialogue systems typically exploit multiple ASR hypothe- ses. … Third, Markov Logic Networks (MLN) have been used to extract slot values by combining probabilistic graphical models and first-order logic [11]. … Related articles

Building Watson: An overview of the DeepQA project [PDF] from aaaipress.org D Ferrucci, E Brown, J Chu-Carroll, J Fan… – AI Magazine, 2010 – aaai.org Page 1. The goals of IBM Research are to advance computer science by exploring new ways for computer technology to affect science, business, and society. Roughly three years ago, IBM Research was looking for a major … Cited by 88 – Related articles – All 25 versions

Automatic recognition of multiparty human interactions using dynamic Bayesian networks [PDF] from ed.ac.uk A Dielmann – 2009 – era.lib.ed.ac.uk … On this challenging re- search domain, we have investigated the use of probabilistic graphical models, in particular dynamic Bayesian networks. … lation and spoken dialogue systems. The proposed features and the multistream DBN infrastructure could be easily … Related articles – All 2 versions

IMPrECISE: Good-is-good-enough Data Integration [PDF] from utwente.nl M van Keulen – 2009 – eprints.eemcs.utwente.nl … We have developed a uniform, closed framework for representing and querying uncertain data based on concepts from probabilistic graphical models; I will present an overview of our framework, the … Contextual Information is essential for adaptive intelligent dialogue systems. … All 5 versions

BibTeX-Entry [PDF] from dagstuhl.de C Koch… – 2009 – drops.dagstuhl.de … We have developed a uniform, closed framework for representing and querying uncertain data based on concepts from probabilistic graphical models; I will present an overview of our framework, the … Contextual Information is essential for adaptive intelligent dialogue systems. … All 2 versions

Personalizing influence diagrams: applying online learning strategies to dialogue management [PDF] from psu.edu DM Chickering… – User Modeling and User-Adapted Interaction, 2007 – Springer … Probabilistic graphical models, such as Bayes- ian networks, have been successfully applied to infer user goals in several domains (eg, Albrecht et … a baseline inffuence diagram for making decisions about what action to take in each step of the dialogue system described above … Cited by 12 – Related articles – BL Direct – All 8 versions

[BOOK] Reinforcement Learning: State-Of-The-Art M Wiering… – 2012 – books.google.com … Application domains range from robotics and computer games to network routing and natural language dialogue systems and reinforcement learning … Advances in the ?eld of probabilistic graphical models are used virtually ev- erywhere, and results for these models-both …

AI’s 10 to Watch [PDF] from stanford.edu J Hendler, P Cimiano, D Dolgov… – Intelligent Systems, …, 2008 – ieeexplore.ieee.org … Such an ontol- ogy-based approach to language processing has important applications not only in ques- tion answering, information extraction, and dialogue systems but also in the Semantic Web, which requires that today’s Web con- tent is formalized using ontologies. … Cited by 1 – Related articles – All 16 versions

Exploring Multimodal Input Fusion Strategies A D’Ulizia – … human computer interaction and pervasive services, 2009 – books.google.com … geographical maps. The Smartkom (Wahlster et al. 2001) is another multimodal dialogue system that merges gesture, speech and facial expressions for both input and output via an anthropomorphic and affective user interface. In the … Related articles – All 5 versions

Email “Speech Acts” V Carvalho – Modeling Intention in Email, 2011 – Springer … Searle [1969], as well as from a number of act taxonomies proposed in the research areas of dialog systems, speech recognition … Dependency networks are probabilistic graphical models in which the full joint distribution of the network is approximated with a set of conditional …

Transformation of graphical models to support knowledge transfer [PDF] from uni-siegen.de A Holland – 2008 – dokumentix.ub.uni-siegen.de … 9 2.2.2. Knowledge Representation . . . . . 11 2.2.3. Knowledge Modeling . . . . . 12 2.2.4. Dialogue System . . . . . 13 2.3. KBS Review . . . . . 15 3. Uncertainty Management 17 3.1. … Related articles – Library Search – All 4 versions

A generic framework for the inference of user states in human computer interaction S Scherer, M Glodek, G Layher, M Schels… – Journal on Multimodal …, 2012 – Springer Page 1. J Multimodal User Interfaces DOI 10.1007/s12193-012-0093-9 ORIGINAL PAPER A generic framework for the inference of user states in human computer interaction How patterns of low level behavioral cues support complex user states in HCI …

[BOOK] Modeling Intention in Email: Speech Acts, Information Leaks and Recommendation Models VR Carvalho – 2011 – books.google.com … This taxonomy drew inspiration from Speech Act Theory Austin [1962], Searle [1969], as well as from a number of act taxonomies proposed in the research areas of dialog systems, speech recognition and machine translation Levin et al. [2003], Paul et al. [1998], Stolcke et al. … Related articles – Library Search – All 2 versions

Linguistic Structure Prediction [PDF] from cmu.edu NA Smith – Synthesis Lectures on Human Language …, 2011 – morganclaypool.com … Semantic Role Labeling Martha Palmer, Daniel Gildea, and Nianwen Xue 2010 Spoken Dialogue Systems Kristiina Jokinen and Michael McTear 2009 Introduction to Chinese Natural Language Processing Kam-Fai Wong, Wenjie Li, Ruifeng Xu, and Zheng-sheng Zhang 2009 … Cited by 5 – Related articles – Library Search – All 6 versions

Incorporating knowledge sources into a statistical acoustic model for spoken language communication systems S Sakti, K Markov… – … , IEEE Transactions on, 2007 – ieeexplore.ieee.org … The matter of commu- nication through speech between human beings and information-processing machines such as dialog systems has also become increasingly important [1]. One of the fundamental technologies for achieving a speech-oriented interface is automatic speech … Cited by 4 – Related articles – BL Direct – All 6 versions

Spatial role labeling: Towards extraction of spatial relations from natural language [PDF] from kuleuven.be P Kordjamshidi, M Van Otterlo… – ACM Transactions on …, 2011 – dl.acm.org … Section 4.4 reports which algorithms, based on probabilistic graphical models, are employed by both subproblems. … In these models, a sentence is a sequence of observations (ie, words), <w1,…,wN>, which can be repre- sented using a probabilistic graphical model. … Cited by 2 – Related articles – All 3 versions

Active affective state detection and user assistance with dynamic Bayesian networks [PDF] from v2.nl X Li… – Systems, Man and Cybernetics, Part A: Systems …, 2005 – ieeexplore.ieee.org … In the READY system [1], the authors use DBNs in a dialog system to adjust the policy in providing instructions, based on the … BNs are probabilistic graphical models representing joint probabilities of a set of random variables and their condi- tional independence relations [19]. … Cited by 74 – Related articles – All 25 versions

[PDF] Towards a separation of pragmatic knowledge and contextual information [PDF] from ceur-ws.org R Porzel, HP Zorn, B Loos… – Proceedings of ECAI-06 …, 2006 – ceur-ws.org … As stated above in a mobile dialogue system contextual informa- tion is of paramount importance as the user expects the offer of … the framework proposed by Porzel [19] can be employed to integrate the various contextual observations in probabilistic graphical models while keep … Cited by 12 – Related articles – View as HTML – All 13 versions

Learning to understand spatial language for robotic navigation and mobile manipulation [PDF] from mit.edu N Roy, TTF Kollar – 2011 – dspace.mit.edu … semantic map of the environment. By extracting a flat sequence of SDCs, we are able to ground the language by using a probabilistic graphical model that is factored into three key components. First, a landmark component … All 2 versions

[PDF] Dynamic dependence analysis: modeling and inference of changing dependence among multiple time-series [PDF] from mit.edu MR Siracusa – 2009 – people.csail.mit.edu Page 1. Dynamic Dependence Analysis : Modeling and Inference of Changing Dependence Among Multiple Time-Series by Michael Richard Siracusa Submitted to the Department of Electrical Engineering and Computer Science in … Related articles – View as HTML – Library Search – All 4 versions

Adapting to the User M Jöst – Handbook of Research on Ubiquitous Computing …, 2007 – books.google.com … Bayesian networks: One of the more prominent approaches dealing with uncer- tainty is Bayesian networks also named Belief networks. These methods can be described as probabilistic graphical models, as they combine probability theory and graph theory. … Cited by 2 – Related articles – All 6 versions

Spotting laughter in natural multiparty conversations: A comparison of automatic online and offline approaches using audiovisual data S Scherer, M Glodek, F Schwenker… – ACM Transactions on …, 2012 – dl.acm.org Page 1. 4 Spotting Laughter in Natural Multiparty Conversations: A Comparison of Automatic Online and Offline Approaches Using Audiovisual Data STEFAN SCHERER, Trinity College Dublin; Ulm University MICHAEL GLODEK …

LEARNING AND REASONING TO ENHANCE ENERGY EFFICIENCY IN TRANSPORTATION SYSTEMS EJ Horvitz… – US Patent App. 11/733,701, 2007 – Google Patents … Sensed Historical data information 1128 User- Computer route dialog system Charging station locations *” Traffic sensin / / Traffic forecasts Commerce opportunities Expected utility optimizer Multiattribute utility model Preferences Distance Destination Speed* energy system … Cited by 3 – Related articles – All 2 versions

[PDF] Error handling in multimodal voice-enabled interfaces of tour-guide robots using graphical models [PDF] from epfl.ch P Prodanov – 2006 – biblion.epfl.ch … 27 3.2 Dialogue-based methods for handling speech recognition errors . . . . . 28 3.2.1 Detecting errors in spoken dialogue systems . . . . . … 32 3.4 Error handling in dialogue systems of service robots . . . . . … Related articles – All 4 versions

[PDF] A hybrid HMM/BN acoustic model utilizing pentaphone-context dependency [PDF] from u-aizu.ac.jp S Sakti, K Markov… – IEICE transactions on …, 2006 – web-ext.u-aizu.ac.jp Page 1. 954 IEICE TRANS. INF. & SYST., VOL.E89-D, NO.3 MARCH 2006 PAPER Special Section on Statistical Modeling for Speech Processing A Hybrid HMM/BN Acoustic Model Utilizing Pentaphone-Context Dependency … Cited by 4 – Related articles – All 9 versions

[CITATION] Distinguished Lecturers, Speech TC, and New Fellows JHL HANSEN – IEEE SIGNAL PROCESSING MAGAZINE, 2005 – ieeexplore.ieee.org … analysis and detection); spoken document retrieval (classification and speech recognition); in-vehicle voice dialog systems and microphone … Michael I. Jordan, Berkeley, Califor- nia: For contributions to probabilistic graphical models and neural information processing systems. …

Learning, logic, and probability: A unified view P Domingos – Lecture Notes in Computer Science, 2004 – books.google.com … In this talk I will describe Markov logic, a representation that combines the full power of first-order logic and probabilistic graphical models, and algorithms for learning and inference in it. … 3068: E. Andre, L. Dybkjser, W. Minker, P. Heis- terkamp (Eds.), Affective Dialogue Systems. … Cited by 1 – Related articles – BL Direct – All 4 versions

[PDF] Using hybrid HMM/BN acoustic models: Design and implementation issues [PDF] from u-aizu.ac.jp K Markov… – IEICE transactions on information and …, 2006 – web-ext.u-aizu.ac.jp Page 1. IEICE TRANS. INF. & SYST., VOL.E89-D, NO.3 MARCH 2006 981 PAPER Special Section on Statistical Modeling for Speech Processing Using Hybrid HMM/BN Acoustic Models: Design and Implementation Issues … Cited by 3 – Related articles – All 9 versions

Understanding gestures with systematic variations in movement dynamics [PDF] from mcgill.ca SCW Ong, S Ranganath… – Pattern recognition, 2006 – Elsevier … Furthermore, a successful sign language recognition system would be useful in many applications such as a sign-to-text/speech translation system, dialog systems for use in specific transactional domains such as government offices and cafeterias, bandwidth-conserving … Cited by 18 – Related articles – All 8 versions

[PDF] MediaHub: An Intelligent MultiMedia Distributed Platform Hub [PDF] from paulmckevitt.com GG Campbell, T Lunney… – 2005 – paulmckevitt.com … model. 2.3.9 SmartKom SmartKom (Wahlster 2003, Wahlster et al. 2001, SmartKom 2005) is a multimodal dialogue system that is being developed to help overcome the problems of interaction between people and machines. The … Related articles – View as HTML – All 8 versions

[PDF] Modeling Intention in Email [PDF] from cmu.edu VR Carvalho – 2008 – lti.cs.cmu.edu Page 1. Modeling Intention in Email Vitor R. Carvalho CMU-LTI-08-007 Language Technologies Institute School of Computer Science Carnegie Mellon University 5000 Forbes Ave., Pittsburgh, PA 15213 www.lti.cs.cmu.edu … Related articles – View as HTML – Library Search – All 9 versions

[PDF] Methods for Life-long Planning and Failure Detection in Dynamic Environments [PDF] from 212.235.189.20 ALU Freiburg – 2006 – 212.235.189.20 … spatial relationships. There are many plan-based dialogue systems that are used (or potentially usable) for HRI (eg, [16; 1]). Most such sys- tems try to exploit the context of the current (dialogue) plan to interpret utterances. In … Related articles – View as HTML – All 5 versions

[BOOK] Managing knowledge in a world of networks: 15th international conference, EKAW 2006, Podebrady, Czech Republic, October 2-6, 2006: proceedings S Staab… – 2006 – books.google.com … A Unified View 2 Domingos 3dge Acquisition .s: Knowledge Acquisition with Repertory Grids and Formal Analysis for Dialog System Construction 3 … talk I will describe Markov logic, a representation that com- es first-order logic and probabilistic graphical models, and algorithms … Library Search – All 2 versions

Data-driven augmentation of pronunciation dictionaries [TXT] from sun.ac.za L Loots – 2010 – scholar.sun.ac.za … The applications of speech synthesis are manifold. Possibly the most widely used applica- tion at present is in automated dialogue systems such as those found in customer call centres. … An HMM is a type of probabilistic graphical model. It is a network with edges representing … Related articles – All 5 versions

[PDF] ICME 2006 [PDF] from securecms.com H TORONTO – 2006 – securecms.com Page 1. ICME 2006 Program Guide INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO JULY 9 – 12, 2006 HILTON TORONTO TORONTO, ONTARIO, CANADA SPONSORED BY THE INSTITUTE OF ELECTRICAL … View as HTML – All 5 versions