Text Classification & Dialog Systems 2014


Notes:

In general, automatic text classification plays a vital role in text summarization, question answering and information extraction.  Automatic text classification is usually done by using a prelabeled training set and applying various machine learning methods such as naive Bayes, support vector machines, artificial neural networks, or hybrid approaches that combine various machine learning methods to improve the efficiency of classification.

  • Automatic text classification
  • Term selection
  • Text categorization
  • Text classifier

Resources:

Wikipedia:

References:

See also:

100 Best GitHub: Text Segmentation | NPCEditor | Stanford ClassifierStatistical Classification & Dialog SystemsText Segmentation 2013 | Text Segmentation 2014


[BOOK] Speech and language processing D Jurafsky, JH Martin – 2014 – cs.colorado.edu … We call programs like HAL that converse with humans in natural language conversational agents or dialogue systems. … A similar spoken dialogue system has been deployed by as- tronauts on the International Space Station. … Cited by 26 Related articles All 20 versions

Biomedical text mining: State-of-the-art, open problems and future challenges A Holzinger, J Schantl, M Schroettner, C Seifert… – … Discovery and Data …, 2014 – Springer … 1070–1087 (2008) CrossRef; Yeh, JF, Wu, CH, Chen, MJ: Ontology-based speech act identification in a bilingual dialog system using partial … 46(1-3), 389–422 (2002) CrossRef; Ghanem, M., Guo, Y., Lodhi, H., Zhang, Y.: Automatic scientific text classification using local patterns … Cited by 11 Related articles All 5 versions

Learning phrase patterns for text classification using a knowledge graph and unlabeled data A Marin, R Holenstein, R Sarikaya… – … Conference of the …, 2014 – mazsola.iit.uni-miskolc.hu … Text classification systems often use lexical features, such as n-grams, and lexicon containment features (ie binary features, holding positive value if … tasks, such as topic clas- sification, more complex problems like intent classification or slot-filling for dialog systems often require … Cited by 4 Related articles All 10 versions

Iterative perceptual learning for social behavior synthesis I de Kok, R Poppe, D Heylen – Journal on multimodal user interfaces, 2014 – Springer Page 1. J Multimodal User Interfaces (2014) 8:231–241 DOI 10.1007/s12193-013- 0132-1 ORIGINAL PAPER Iterative perceptual learning for social behavior synthesis Iwan de Kok · Ronald Poppe · Dirk Heylen Received: 13 … Cited by 6 Related articles All 14 versions

Learning situated knowledge bases through dialog A Pappu, AI Rudnicky – Proceedings of the 15th …, 2014 – mazsola.iit.uni-miskolc.hu … [4] P. Wang and C. Domeniconi, “Building semantic kernels for text classification using wikipedia,” in … [6] A. Pappu and A. Rudnicky, “Predicting tasks in goal-oriented spo- ken dialog systems using semantic knowledge bases,” in Proceed- ings of the SIGDIAL 2013 Conference. … Cited by 3 Related articles All 5 versions

Comparing multi-label classification with reinforcement learning for summarisation of timeseries data D Gkatzia, H Hastie, O Lemon – 52nd Annual Meeting of the …, 2014 – macs.hw.ac.uk … In contrast, ML clas- sification does not need the computation of links between the data and the templates. ML classi- fication can also apply to other problems whose features are correlated, such as text classification (Madjarov et al., 2012), when an aligned dataset is provided. … Cited by 5 Related articles All 10 versions

An impact analysis of features in a classification approach to irony detection in product reviews K Buschmeier, P Cimiano, R Klinger – ACL 2014, 2014 – anthology.aclweb.org … The second baseline relies on standard text classification features that are not specific for the task. … 2006.“yeah right”: Sarcasm recogni- tion for spoken dialogue systems. In Proceedings of InterSpeech, pages 1838–1841, Pittsburgh, PA, September. … Cited by 4 Related articles All 11 versions

Wikipedia-based Kernels for Dialogue Topic Tracking S Kim, RE Banchs, H Li – Proceedings of ICASSP, 2014 – rbanchs.com … Recently, some researchers [12, 13] have shown the feasibility of using Wikipedia knowledge to build dialogue systems. … According to the previous work on Wikipedia-based semantic kernels for text classification [14], an extended feature space is defined by concatenating the … Cited by 2 Related articles All 3 versions

Automatic detection of psychological distress indicators and severity assessment in crisis hotline conversations M Pacula, T Meltzer, M Crystal… – … , Speech and Signal …, 2014 – ieeexplore.ieee.org … Such rationales proved useful in related text classification tas/s in literature [7], [8]. In addition to distress labels, each chat was also annotated with a triage code indicating treatment acuity … Bayesian update of dialogue state: A POMDP framework for spoken dialogue systems. … Cited by 1 Related articles All 3 versions

Language Technology for eGovernment–Business Cases M Henkel, E Perjons, E Sneiders, J Karlgren… – New Perspectives in …, 2014 – Springer … 284 (2000) Acomb, K., Bloom, J., Dayanidhi, K., Hunter, P., Krogh, P., Levin, E., Pieraccini, R.: Technical support dialog systems: Issues, problems … 25(3), 361–388 (1999) Colas, F., Brazdil, P.: Comparison of SVM and some older classification algorithms in text classification tasks … Cited by 1 Related articles All 5 versions

Associating targets with SentiUnits: a step forward in sentiment analysis of Urdu text AZ Syed, M Aslam, AM Martinez-Enriquez – Artificial Intelligence Review, 2014 – Springer Page 1. Artif Intell Rev (2014) 41:535–561 DOI 10.1007/s10462-012-9322-6 Associating targets with SentiUnits: a step forward in sentiment analysis of Urdu text Afraz Z. Syed · Muhammad Aslam · Ana Maria Martinez-Enriquez … Cited by 3 Related articles All 7 versions

Knowledge Discovery Through Spoken Dialog A Pappu – 2014 – speech.cs.cmu.edu … All processes in a dialog system pipeline, such as recognition, parsing and dialog management, depend on the words in an input. … 1.1 Handling New Information Spoken Dialog Systems are challenged by misunderstandings and non- understanding errors. … Cited by 1 Related articles All 6 versions

Microblogging as a mechanism for human–robot interaction D Bell, T Koulouri, S Lauria, RD Macredie… – Knowledge-Based …, 2014 – Elsevier This paper presents a novel approach to social data analysis, exploring the use of microblogging to manage interaction between humans and robots, and presenting. Cited by 9 Related articles All 6 versions

An efficient gradient-based approach to optimizing average precision through maximal figure-of-merit learning I Kim, CH Lee – Journal of Signal Processing Systems, 2014 – Springer Page 1. J Sign Process Syst (2014) 74:285–295 DOI 10.1007/s11265-013-0748-0 An Efficient Gradient-based Approach to Optimizing Average Precision Through Maximal Figure-of-Merit Learning Ilseo Kim · Chin-Hui Lee Received … Cited by 1 Related articles All 4 versions

Effective hand gesture classification approaches P Premaratne – Human Computer Interaction Using Hand Gestures, 2014 – Springer … SVM has been found attractive by researchers as it is known as a quadratic programming problem with application found in hand gesture recognition [26–32], and text classification methods [33–37]. These applications Fig. 5.11 Cited by 1 Related articles

A hierarchical intention recognition model for situated dialogue system Y Liu, X Wang – … and Intelligence Systems (CCIS), 2014 IEEE …, 2014 – ieeexplore.ieee.org … 2.2 The hierarchical intention recognition model in situated dialogue system … Features are a very important factor in supervised classification, and the features of dialogue are significantly different from those for text classification. …

Bootstrapping Dialog Systems with Word Embeddings G Forgues, J Pineau, JM Larchevêque, R Tremblay – cs.cmu.edu … Abstract One of the main tasks of a dialog system is to assign intents to user utterances, which is a form of text classification. Since intent labels are application-specific, bootstrapping a new dialog system requires collecting and annotating in-domain data. … Related articles

An empirically-based, tutorial dialogue system: design, implementation and evaluation in a first year health sciences course. J McDonald, A Knott, S Stein, R Zeng – ascilite.org.au … results reported here are the first from what we hope will be many in-class studies conducted using this tutorial dialogue system and others like … There is room to improve the language recognition or text classification part of the system and perhaps this alone may result in greater … Related articles All 4 versions

Topic classification of spoken inquiries using transductive support vector machine R Torres, H Kawanami, T Matsui, H Saruwatari… – Natural Interaction with …, 2014 – Springer … J. Mariani et al. (eds.), Natural Interaction with Robots, Knowbots and Smartphones: Putting Spoken Dialog Systems into Practice, DOI 10.1007/978-1-4614-8280-2 23 … TSVMs were proposed by Vapnik in 1998 and were introduced by Joachims [2] for text classification. … Related articles All 5 versions

Text Mining Practical: Project Ideas B QasemiZadeh, S Handschuh – atmykitchen.info … As a project title in the domain of text classification, we suggest the task of short text classification, which is more challenging … The research in natural language dialogue systems aims to develop methods and models that can understand and generate linguistic expressions, in …

Detecting Multiple Domains from User’s Utterance in Spoken Dialog System S Ryu, J Song, S Koo, S Kwon, GG Lee – uni-ulm.de … S. Han, GG Lee, M. Kim, and K. Kim, “A Hierarchical Domain Model-Based Multi-Domain Selection Framework for Multi-Domain Dialog Systems,” In Proceedings of … [6] A. McCallum and K. Nigam, “A comparison of event models for Naive Bayes text classification,” In Proceedings … Cited by 1 Related articles All 2 versions

User Input Classification for Chinese Question Answering System Y Hou, X Wang, Q Chen, M Li, C Tan – Machine Learning and Cybernetics, 2014 – Springer … some users think the QA system knows everything and input queries from all domains, or use the QA system as a dialogue system and send … The features were selected based on Chinese short text classification methods [9] and the special characters for the QA system user input … Related articles All 3 versions

Satire Detection from Web Documents using machine Learning Methods T Ahmad, H Akhtar, A Chopra… – Soft Computing and …, 2014 – ieeexplore.ieee.org … The difficulty in satire detection results in misunderstanding in communication and also pose problem for NLP systems like dialogue system, online review systems … Text classification is the task of assigning any kind of topic category to any piece of text and that could be subject … Related articles All 2 versions

A Composite Kernel Approach for Dialog Topic Tracking with Structured Domain Knowledge from Wikipedia S Kim, RE Banchs, H Li – anthology.aclweb.org … in conversational contexts. This fact sug- gests that a dialog system should be also capable of conducting multi-topic conversations with users to provide them a more natural interaction with the system. However, the majority … Cited by 1 Related articles All 7 versions

Exploiting out-of-vocabulary words for out-of-domain detection in dialog systems S Ryu, D Lee, GG Lee, K Kim… – Big Data and Smart …, 2014 – ieeexplore.ieee.org … For example, the HDM of a Smart TV dialog system have three base domains: TV Program Guide, Video-on-Demand (VOD) Guide, and … 1 Some researchers argue that these three domains can be designed as just one domain; this makes sense in text-classification perspective. … Related articles All 3 versions

Estimating the Sentiment of Arabic Social Media F Harrag – citala.org … [12] Hijjawi M. and Bander Z., “An Arabic Stemming Approach using Machine Learning with Arabic Dialogue System” in ICGST AIML-11 Conference, Dubai, UAE, 12-14 April 2011. … [16] Khreisa L., “Arabic Text Classification Using N-Gram Frequency Statistics Comparative Study … Related articles

A pragmatic tutorial dialogue system: design, implementation and evaluation in a health sciences domain JA McDonald – 2014 – ourarchive.otago.ac.nz … 32 3.2.3.1 Dialogue system architecture . . . . . 32 3.2.3.2 Dialogue management models . . . . . 33 3.2.3.3 Natural language understanding . . . . . 36 3.3 Text classification . . . . . … Related articles All 2 versions

Experiences with an Approach to Abstract Handling of Content for Human Machine Interaction Applications R Schmidt, J Fonfara, S Hellbach, HJ Böhme – ceur-ws.org … Thirdly, having all the user utterances annotated, we tested several algo- rithms for text classification. … But to further improve this use case, a simple dialog system could be added, that employs data from our people tracker to allow the robot to react to nearby persons and … Related articles

A Complete Bibliography of ACM Transactions on Speech and Language Processing (TSLP) NHF Beebe – 2014 – tug.ctan.org … Higashinaka:2004:EDU [HMNA04] Ryuichiro Higashinaka, Noboru Miyazaki, Mikio Nakano, and Kiyoaki Aikawa. Evaluating dis- course understanding in spoken dialogue systems. ACM Trans- actions on Speech and Language Processing (TSLP), 1(1):1–20, November 2004. … Related articles All 10 versions

Improving speech-based PTSD detection via multi-view learning X Zhuang, V Rozgic, M Crystal… – … Workshop (SLT), 2014 …, 2014 – ieeexplore.ieee.org … Index Terms— PTSD screening, speech recognition, text classification, audio analysis, EEG signal processing, multi- view learning … including heart rate [5], heart rate and galvanic skin response (GSR) [6], EEG [7], voice qual- ity [8], speech [9], speech in dialog systems [10, 11 … Related articles

Task Estimation Using Latent Semantic Analysis of Visual Scenes and Spoken Words M Kimura, S Sawada, Y Iribe… – Electronics and …, 2014 – Wiley Online Library … For example, it has been pointed out in the field of text classification that an advanced classification capability can be achieved by using support vector machines and other algorithms for word categorization in a vector space [18]. REFERENCES. … Related articles All 3 versions

CLARA: a multifunctional virtual agent for conference support and touristic information LF D’Haro, S Kim, KH Yeo, R Jiang, AI Niculescu… – researchgate.net … Here, the module in- cludes a text classifier that is able to detect the topic of search. … Biomedical Informatics, 2006 Oct;39(5):482-99 [2] Maragoudakis, M. MeteoBayes: Effective plan recognition in a weather dialogue system. IEEE Intelligent Systems, 22(1), p. 66-77. … Cited by 2 Related articles All 4 versions

Ontology Based Contextual Intent Recognition for Speech Gesture Utterances MR Saktheeswari, MN Devi – ijritcc.org … Word Net knowledge into text representation that can lead to significant reductions in error rates on certain types of text classification tasks Using … an important part of a dialog system, we are planning to learn new user contexts, which can be represented in the ontology by the … Related articles All 3 versions

Adaptive operator switching and solution space probability structure based genetic algorithm for information retrieval through pattern recognition J Dalai, SZ Hasan, B Sarkar… – Circuit, Power and …, 2014 – ieeexplore.ieee.org … Dialogue system also came into picture during that decade itself. A variety of advanced techniques for text based information retrieval trough web search are available, which include efficient text indexing, Boolean and vector space retrieval models, text classification and … Related articles

Semantic Similarity Calculation of Short Texts Based on Language Network and Word Semantic Information Z Zhan, F Lin, X Yang – Advanced Computer Architecture, 2014 – Springer … it can extend the inquiry interface [4]. Moreover, it also has important applications in health advisory dialogue system [5], property … We choose experiment data from the partial text classification corpus library gathered and organized by the natural language processing group … Related articles All 3 versions

Hot Topics and Schisms in NLP: Community and Trend Analysis with Saffron on ACL and LREC Proceedings P Buitelaar, G Bordea… – 9th Edition of Language …, 2014 – ir.library.nuigalway.ie … and Romary, 2010), information extrac- tion (Yangarber et al., 2000), domain ontology construction (Kietz et al., 2000), text classification (Basili et al … in the top-left part of Figure 4. This top- ical hierarchy identifies the terms ”speech synthesis“ and ”dialogue systems“ as subtopics … Related articles All 4 versions

Report on learning semantic analysis components A Vlachos, S Clark – 2014 – spacebook-project.eu … Unlike in typical text classification tasks, content words are not always helpful in dialog act tagging; eg the word “church” in the user utterances in Figure 1 is not indicative of set question, while n-grams of words typically considered as stopwords, such as “what is this”, can be … Related articles All 2 versions

Civilian Analogs of Army Tasks: Supporting Pedagogical Storytelling Across Domains A Gordon, M Core, SH Kang, C Wang, C Wienberg – ict.usc.edu … Using supervised machine learning techniques, Swanson constructed a text classifier to identify these personal stories in streams of weblog posts (precision 59.1 … Using a menu-based dialogue system, learners select conversational moves in an evolving dialogue context. … Related articles All 4 versions

A Shared, Modular Architecture for Developing Virtual Humans A Hartholt, D Traum, S Marsella, LP Morency… – ict.usc.edu … The Agent, for example, can range from rule-based systems [5], to a statistical text-classifier (the NPCEditor [13] which also serves … Sagae K., Gerten J., Nazarian A., Traum D., FLoReS: A Forward Looking, Reward Seeking, Dialogue Manag- er, Spoken Dialog Systems (2012) 16 … Related articles

Automated Indexing of Internet Stories for Health Behavior Change: Weight Loss Attitude Pilot Study R Manuvinakurike, WF Velicer… – Journal of medical …, 2014 – ncbi.nlm.nih.gov … of change that are most useful for each stage of change for weight loss yields only 5 unique sets of processes that need to be identified, decreasing the complexity of the text classification problem from … Bickmore T, Giorgino T. Health dialog systems for patients and consumers. … Related articles All 8 versions

Preliminary Research and Development on the Problem-Solving Humanoid Machine for K-12 Education X Yu, D Yang, W Ding, W Gan – Educational Innovation through …, 2014 – ieeexplore.ieee.org … testing the related technologies such as text to speech, speech recognition, document understanding, formula recognition, text classification, problem-solving … Dialogue systems have employed text, speech, face, haptic, gestures and other modes to communicate with students. … Related articles All 2 versions

A New Corpus and Imitation Learning Framework for Context-Dependent Semantic Parsing A Vlachos, S Clark – Transactions of the Association for …, 2014 – tacl2013.cs.columbia.edu … (h) Atomic predictions Figure 2: Semantic parsing decomposition. bigrams and trigrams and the final punctuation mark. Unlike in typical text classification tasks, content words are not always helpful in dialog act tagging; eg the token “meal” in Fig. … Cited by 1 Related articles All 6 versions

Application of deep belief networks for natural language understanding R Sarikaya, GE Hinton, A Deoras – Audio, Speech, and …, 2014 – ieeexplore.ieee.org … fine-tuned with backpropagation. We compare a DBN-initialized neural network to three widely used text classification algorithms: support vector machines (SVM), boosting and maximum entropy (MaxEnt). The plain DBN-based … Cited by 20 Related articles All 8 versions

Interacting with Traditional Chinese Culture through Natural Language X Wang, ET Khoo, R Nakatsu, A Cheok – Journal on Computing and …, 2014 – dl.acm.org … Systems—Natural language interfaces General Terms: Algorithms, Design Additional Key Words and Phrases: Conversational agents, dialogue systems, intangible cultural … 2010]. The system works by having a statistical text classification algorithm to map the utterance text to … Related articles

Analysis of word co-occurrence in human literature for supporting semantic correspondence discovery J Martinez-Gil, M Pichler – … of the 14th International Conference on …, 2014 – dl.acm.org … Past works in this field include the automatic processing of text and email messages [15], healthcare dialogue systems [5], natu- ral … sources [17], c) data mining where using appropriate semantic similarity measures can facilitate the processes of text classification and pattern … Related articles

Latent semantic rational kernels for topic spotting on conversational speech C Weng, DL Thomson, P Haffner… – Audio, Speech, and …, 2014 – ieeexplore.ieee.org … K(A, B) = 0@ M (x,y)2S??[ATB](x, y) 1 A = (w[ATB]) (5) B. N-gram Rational Kernels An N-gram kernel is a rational kernel that has PDS property [7] and has been widely and successfully used in speech or text classification applications [13]. … Related articles All 2 versions

An introduction to question answering over linked data C Unger, A Freitas, P Cimiano – Reasoning Web. Reasoning on the Web …, 2014 – Springer Page 1. An Introduction to Question Answering over Linked Data Christina Unger 1 , André Freitas 2 , and Philipp Cimiano 1 1 CITEC, Bielefeld University, Inspiration 1, 33615 Bielefeld 2 Insight, National University of Galway (NUIG), Galway Abstract. … Cited by 4 Related articles All 4 versions

Automatic generation of multiple choice questions using dependency-based semantic relations N Afzal, R Mitkov – Soft Computing, 2014 – Springer … (2011). 3.2 Question generation Automatic question generation is an important and emerg- ing area of research in NLP. Automatic question generation has the potential to be employed in various areas such as intelligent tutoring systems, dialogue systems (Walker et al. … Cited by 3 Related articles All 4 versions

[BOOK] Text Mining of Web-based Medical Content A Neustein – 2014 – books.google.com … The chapter on user satisfaction with a health dialogue system is followed by a fascinating presentation of the Smith-Kettlewell Eye Research … 2 Extraction of features and semantic information 15 1.3.3 Analysis of annotated texts 16 1.3.3.1 Algorithms for text classification 18 … Related articles All 4 versions

[BOOK] Natural Language Processing with Java and LingPipe Cookbook B Baldwin, K Dayanidhi – 2014 – books.google.com … Since first being introducedto the fieldin 2006,hehas worked ondiverse problems suchas spoken dialog systems, machine translation, text normalization, coreference resolution, and speechbased information retrieval, leadingto publications in esteemed conferences suchas … All 2 versions

Agile Facilitation for Collaborative Learning D Adamson – 2014 – cs.cmu.edu … Using discourse analysis methods, sequential modeling, and text classification, we will explore the potential of modeling conversational context and trajectory to select the most appropriate strategies for collaborative facilitation. Page 6. February 19, 2014 DRAFT vi Page 7. … Related articles All 2 versions

Part-of-speech tagging in written slang V Korolainen – 2014 – jyx.jyu.fi … LIST OF FIGURES FIGURE 1: Architecture pipeline for a Spoken Dialogue System….. … FIGURE 1 shows that a spoken dialogue system have nine different stages. When each of the stages have multiple subtasks, it is not surprising that the total process might take time. … Related articles

[BOOK] An Empathic Virtual Buddy for Social Support JM Van der Zwaan – 2014 – repository.tudelft.nl Page 1. Page 2. Page 3. An Empathic Virtual Buddy for Social Support Page 4. Page 5. An Empathic Virtual Buddy for Social Support Proefschrift ter verkrijging van de graad van doctor aan de Technische Universiteit Delft, op gezag van de Rector Magnificus prof. ir. … Cited by 1 Related articles All 2 versions

Foundations and Trends in Signal Processing L Deng, Y Dong – Signal Processing, 2014 – research.microsoft.com Page 1. the essence of knowledge FnT SIG 7:3-4 Deep Learning; Methods and Applications Li Deng and Dong Y u Foundations and Trends® in Signal Processing 7:3-4 Deep Learning Methods and Applications Li Deng and Dong Yu now now Page 2. 7.1. … Cited by 3 Related articles All 12 versions

Combining visual recognition and computational linguistics: linguistic knowledge for visual recognition and natural language descriptions of visual content M Rohrbach – 2014 – scidok.sulb.uni-saarland.de Page 1. Combining Visual Recognition and Computational Linguistics Linguistic Knowledge for Visual Recognition and Natural Language Descriptions of Visual Content Thesis for obtaining the title of Doctor of Engineering Science (Dr.-Ing.) … Related articles All 4 versions