Maxent (Maximum Entropy Modeling Toolkit) 2011

Maximum Entropy Modeling Toolkit for Python and C++ github: maxent {related:}


Principle of maximum entropyMaximum entropy models

[PDF] The Karlsruhe Institute of Technology translation systems for the WMT 2011 [PDF] from T Herrmann, M Mediani, J Niehues… – Proceedings of the Sixth …, 2011 – … Precision Recall F-measure Regression 93.81 87.27 90.42 LogReg 93.43 94.84 94.13 MaxEnt 93.69 94.54 … The maximum entropy classifier performed better than the logistic regression in terms of precision but however it … SRILM – An Extensible Lan- guage Modeling Toolkit. … Cited by 3 – Related articles – View as HTML – All 9 versions

Emotion recognition of affective speech based on multiple classifiers using acoustic-prosodic information and semantic labels C Wu, W Liang – Affective Computing, IEEE Transactions on, 2011 – … If these two parameters are greater than the threshold value, the union of A and B will be retained for Maximum Entropy (MaxEnt) modeling. To model the emotion based on semantic labels, MaxEnt is employed to model the abovementioned EARs as follows. … Cited by 6 – Related articles – All 4 versions

[PDF] ILLC-UvA translation system for EMNLP-WMT 2011 [PDF] from M Khalilov… – Proceedings of the Sixth Workshop on Statistical …, 2011 – … 5.2 Experimental setup Moses toolkit (Koehn et al., 2007) in its standard setting was used to build the SMT systems: • GIZA++/mkcls (Och, 2003; Och, 1999) for word alignment. … For maximum entropy modeling we used the maxent toolkit4. … Cited by 1 – Related articles – View as HTML – All 9 versions

[PDF] Enriching text-to-speech synthesis using automatic dialog act tags [PDF] from VKR Sridhar, A Syrdal, A Conkie… – Proceedings of …, 2011 – … We use the machine learning toolkit LLAMA [7] to estimate the conditional distribution using maxent. 2.2. Training Data for Dialog Act Tagger … Table 1: Dialog act tagging accuracies (in %) for lexical and syntactic cues with the maximum entropy model. … Cited by 1 – Related articles – All 4 versions

MAXIMUM-ENTROPIE-METHODEN IN MUSTERERKENNUNG UND K UNSTLICHER INTELLIGENZ EG Schukat-Talamazzini – … 3] 10. Das MaxEnt-Toolkit von Ristad Ris98, Laf95] 1 … Luc9X] H. Lucke: The Edge Pruning Algorithm for Finding Decomposable Stochastic Models, IEEE Trans. on Pattern Analysis and Machine Intelligence, 199X. Ris98] ES Ristad: Maximum Entropy Modeling Toolkit, 1998. … Related articles – View as HTML – All 2 versions

A Maximum Entropy Segmentation Model for Statistical Machine Translation D Xiong, M Zhang… – Audio, Speech, and Language Processing, … – … is constructed with cohesive segment boundary classifiers which are based on the maximum entropy principle. … all corpora except for the United Nations corpus to train our MaxEnt based reordering … We built a four-gram language model using the SRILM toolkit [21], which was … Related articles

Chinese event place phrase recognition of emergency event using Maximum Entropy F Zhu, Z Liu, J Yang… – Cloud Computing and …, 2011 – … [10] Le Zhang, “Maximum Entropy Modeling Toolkit for Python and C++,” http://homepages.inf. _toolkit.html. [11] AL Berger, et al., “A maximum entropy approach to natural language processing,” Computational linguistics, vol. 22, pp. 39-71, 1996. 618 Related articles

Cross-lingual sentiment classification via bi-view non-negative matrix tri-factorization J Pan, GR Xue, Y Yu… – … in Knowledge Discovery and Data Mining, 2011 – Springer … 6 7 shtml 8 9 We use Maximum Entropy Modeling Toolkit for Python and C++ by Le Zhang, toolkit.html … Related articles – All 2 versions

[PDF] CREATING ENSEMBLE OF DIVERSE MAXIMUM ENTROPY MODELS [PDF] from K Audhkhasi, A Sethy, B Ramabhadran… – … principle aims to find a proba- bility distribution P(y|x) with maximum entropy subject to … method (such as L- BFGS [12]) can be used to find the maximum likelihood parameter … The ?MaxEnt model ensemble performs significantly better than 1 MaxEnt model trained on the entire … View as HTML

Using Rich Linguistic and Contextual Information for Tree-Based Statistical Machine Translation BT Hung, N Le Minh… – … on Asian Language …, 2011 – … Under the MaxEnt model, we have: … Meanwhile, we gather the lexical and syntax features for training the Maximum entropy models. The maximum initial phrase length is set to 10, and the maximum rule length of the source side is set to 5. We use SRILM toolkit [11] to train …

Detecting structural events for assessing non-native speech [PDF] from L Chen… – Proceedings of the 6th Workshop on Innovative Use …, 2011 – … extracted on human annotations, the features derived from struc- tural events automatically estimated by the two NLP models show a lower but sufficiently high r. The features derived from the MaxEnt model’s estima- tions on … Maximum Entropy Model- ing Toolkit for Python … Related articles – All 13 versions

Comparative Analysis of Tools Available for Developing Statistical Approach Based Machine Translation System A Kumar… – Information Systems for Indian Languages, 2011 – Springer … YASMET [W10]: YASMET stands for Yet Another Small MaxEnt Toolkit. This is a tiny toolkit for performing training of maximum entropy models. … Joshua [W16]: It is an open source decoder for statistical translation models based on synchronous context free grammars. … Related articles – All 2 versions

[PDF] Supervised classification of news articles by whether they mention diseases and outbreaks [PDF] from M Chua – 2011 – … To train a MaxEnt classifier, it is necessary to determine the value of each <feature, class>’s weight based on the training … This algorithm is implemented by an existing library, the Maximum Entropy Modeling Toolkit for Python and C++ 1. My code reads the training corpus, maps … Related articles – View as HTML – All 2 versions

[PDF] ETS: An Error Tolerable System for Coreference Resolution [PDF] from H Xiong, L Song, F Meng, Y Liu, Q Liu… – ACL HLT 2011, 2011 – … 2005. Coarse-to-fine n- best parsing and maxent discriminative reranking. In Proceedings of the 43rd Annual Meeting on Associ- ation for Computational Linguistics, pages 173-180. … Z. Le. 2004. Maximum entropy modeling toolkit for Python and C++. … Related articles – View as HTML – All 11 versions

A Model Lexicalized Hierarchical Reordering for Phrase Based Translation V Van Nguyen, TP Nguyen, M Le Nguyen… – Procedia-Social and …, 2011 – Elsevier … Then we use the open source toolkit for Maximum Entropy 1 to train Maximum Entropy model for reordering model. … To integrate HRM model into decoding, we compute reordering score 1 toolkit.html 82 Vinh Van Nguyen et al. …

[PDF] A Unified and Discriminative Soft Syntactic Constraint Model for Hierarchical Phrase-based Translation [PDF] from LLTZC Wang… – … Page 6. SRILM Toolkits (Stolcke, 2002) with modified Kneser-Ney smoothing (Chen and Goodman,1998). … Table 4.The translation effect of unifying methods for MaxEnt based SSC models on MT NIST05 test set. … 2006. Maximum Entropy Modeling Toolkit for Python and C++. … Related articles – View as HTML

[PS] Use of Maximum Entropy in Back-o Modeling for a Named Entity Tagger [PS] from RK Srihari, CN Munirathnam Srikanth… – … 1 oldc(Y ) c(Y ) ? 1 1 + uniqueoutcomesof Y c(Y ) . Page 5. HKK CONFERENCE 5 VIII. Use of ME for Estimating Name Class: No Backoff In the nal model to be discussed, no backo model is employed. … In particular, Jaynes dis- cusses the maximum entropy principle (MaxEnt). … Related articles – View as HTML

Knowledge transfer based on feature representation mapping for text classification J Meng, H Lin… – Expert Systems with Applications, 2011 – Elsevier … IEEE Transactions on Pattern Analysis and Machine Intelligence, 23(4), 396-403. Zhang, L. Maximum entropy modeling toolkit for python and c++. Available at: http://homepages.inf.ed. Zhang, T., & Oles, F. (2001). … Related articles – All 3 versions

Comparison of five modelling techniques to predict the spatial distribution and abundance of seabirds S Oppel, A Meirinho, I Ramírez, B Gardner… – Biological …, 2011 – Elsevier … ( Maximum entropy is a presence-only modelling approach that uses background samples of the environment rather than absence locations to estimate environmental relationships. …

[PDF] Linking Named Entities to a Structured Knowledge Base [PDF] from VV PINGALI – … IR = Information Retrieval, NB = Naive Bayes, MaxEnt = Maximum Entropy … based approach is able to perform slightly better than Naive Bayes and Maximum Entropy approaches. … Bow: A toolkit for statistical language modeling, text re- trieval, classification and clustering, 1996. … Related articles – View as HTML

[PDF] Using Human Perception for Automatic Accent Assessment [PDF] from F William, A Sangwan… – 2011 – … In this study, the features for the proposed MEM (Maximum Entropy Model) are pronunciation errors, ie … method for alignments between decoded and canonical phones sequences, and Maxent Toolkit [9] to train … After the models are obtained, the testing set is evaluated using Eq. … Related articles

Recognition of word emotion state in sentences C Quan… – IEEJ Transactions on Electrical and …, 2011 – Wiley Online Library … Maximum entropy modeling provides a framework for integrat- ing information from many heterogeneous information sources … The deter- ministic model produces conditional probability [25], see (1) and (2). In experiments, we have used a Java-based open-nlp MaxEnt toolkit.2 … Related articles – All 2 versions

Error detection in broadcast news ASR using Markov chains [PDF] from T Pellegrini… – Human Language Technology. Challenges for …, 2011 – Springer … The HTK toolkit2 was used to train and test MCs … tested by comparing this approach to a Gaussian Mixture Model (GMM), and to a maximum entropy model, that is currently used in our … A Maxent model with a 0.5 decision threshold was able to detect only 7% of the errors correctly … Related articles – All 4 versions

Resolving Noun Phrase Coreference in Czech M Novák… – Anaphora Processing and Applications, 2011 – Springer … Method Precision Recall F-Measure MaxEnt classifier 57.30% 33.54% 42.32% MaxEnt ranker 58.55% 35.58 … From this point of view, maximum entropy ranker performs better than perceptron ranker. Except for the final evaluation we were interested how models deal with quan …

Identification of conjunct verbs in hindi and its effect on parsing accuracy [PDF] from R Begum, K Jindal, A Jain, S Husain… – … and Intelligent Text …, 2011 – Springer … These features are mapped to binary features. We used maximum entropy toolkit2 to conduct our experiments. 6 Experiments and Results … The best model gives us the highest accuracy of around 85.28%. … 2 toolkit.html Page 8. … Cited by 3 – Related articles – All 4 versions

[PDF] A Cross-lingual Annotation Projection-based Self-supervision Approach for Open Information Extraction [PDF] from S Kim, M Jeong, J Lee… – … using CRF++ and Maximum Entropy Modeling toolkits 7. 4.3 Extraction During execution, the input of the system is raw text in Korean and the output is a set of extractions 6http://sourceforge. net/projects/mstparser/ 7 toolkit.html 745 … Related articles – View as HTML – All 3 versions

[PDF] On Classifying the Political Sentiment of Tweets [PDF] from C Johnson, P Shukla… – … scenario such as this, this is equivalent to fitting a logistic regression model to the … In implement- ing a maximum entropy classifier portion of our project we extended the updown sentiment … Go’ were then resolved to the same com- mon identifier allowing our MaxEnt classifier to … View as HTML

System and method of exploiting prosodic features for dialog act tagging in a discriminative modeling framework S Bangalore… – US Patent 7,996,214, 2011 – Google Patents … to embodiments within the scope of the present invention, a chunk based model for dialog act … at least one parameter in the user utterance by using maximum entropy discriminative classi?ers. … distribution P(d,-l<I>) )by choosing the maxi- mum entropy (maxent) distribution that … Related articles – All 3 versions

[PDF] Context-Sensitive Syntactic Source-Reordering by Statistical Transduction [PDF] from M Khalilov… – … All language models were trained with SRI LM toolkit (Stolcke, 2002). Language models for Dutch, Spanish and Chinese use 5-grams, while the ideal- … For Maximum Entropy model- ing we used the maxent toolkit7. Translation scores. … Related articles – View as HTML – All 4 versions

Named Entity Recognition System for Hindi Language: A Hybrid Approach [PDF] from S Srivastava, M Sanglikar… – International Journal of …, 2011 – … MaxEnt [6]: MaxEnt stands for Maximum Entropy Markov Model (MEMM). … 5.2 MaxEnt Based Machine Learning It is based on the principle of maximum entropy which states that the least biased model which considers all know facts is the one which maximizes entropy. … Related articles – All 4 versions

[PDF] Bootstrapped Named Entity Recognition for Product Attribute Extraction [PDF] from DP Putthividhya… – … Maximum entropy classifier is thus closely related to logistic regression model … is analogous to compar- ing MAP (maximum a priori) estimate with that of ML (maximum likelihood) in … The MaxEnt implementation used in our experiment is the version available from the NLTK toolkit … Related articles – View as HTML – All 5 versions

[PDF] Detecting forum authority claims in online discussions [PDF] from A Marin, B Zhang… – ACL HLT 2011, 2011 – … The classifier used is a maximum entropy clas- sifier (MaxEnt), implemented using the MALLET package (McCallum … The results are summarized in table 4. Maximum# selected Dev Eval n-gram length … one that provides the highest dev set F-score, which is the MaxEnt clas- sifier … Cited by 1 – Related articles – View as HTML – All 13 versions

[PDF] Machine Transcription of Guqin Tablature and Automatic Music Rhythm Tagging? [PDF] from Y Guan… – Journal of Information & Computational Science, 2011 – … di = agrdmax P(di|?i) The maximum entropy model is one of the techniques successfully used in the research on part- of-speech tagging. We employ the Maxent Toolkit [16] with selected features as follows: di; ?i-5 to ?i-1; ?i+1g to g?i+5; di-3di-2di-1. … Related articles – View as HTML

[PDF] Function Word Generation in Statistical Machine Translation Systems [PDF] from L Cui, D Zhang, M Li… – … Given the contextual information C, TFWIM is computed by the maximum entropy (ME) approach as follows: … The proposed TFWIM is trained by the MaxEnt toolkit (Zhang, 2006 … The training data of the lexicalized reordering model comes from LDC2003E14, which contains 128K … Related articles – View as HTML

[PDF] OntoGene at CALBC II and Some Thoughts on the Need of Document-Wide Harmonization [PDF] from S Clematide, F Rinaldi… – Second CALBC Workshop – … 4 core). Therefore, MaxEnt was adopted in our system. In … new entity type. Then we extracted the features mentioned above to get the well-formatted training data for maximum entropy tool kit to train model. Thirdly, we dealt … Related articles – View as HTML – All 2 versions

[PDF] Detection and Classification of Network Intrusion Using ILACR [PDF] from R Ramprasath… – International Journal of Computing … – … The simplest conditional classifier is the Maxent classifier based upon maximum entropy classification, which … In this case, using a conditional model would result in a simple maximum entropy classifier [40 … time efficiency and not in the time required for training of the model as the … Related articles – View as HTML

[PDF] RTextTools: A Supervised Learning Package for Text Classification [PDF] from TP Jurka… – … Each model created in the previous step is passed on to … and the two highest per- forming algorithms (maximum entropy and random … recall_accuracy(analytics@doc_summary$ MANUAL_CODE,analytics@doc_summary$ FORESTS_LABEL) # MAXENT 57% ACCURACY … Related articles – View as HTML

Eco-cultural niches of the Badegoulian: Unraveling links between cultural adaptation and ecology during the Last Glacial Maximum in France WE Banks, T Aubry, F d’Errico, J Zilhão… – Journal of …, 2011 – Elsevier … were produced with the sum function of a uniform phase model in OxCal … Furthermore, the Badegoulian presents a simplified and relatively low-investment toolkit (see Straus … used genetic algorithm (GARP; Stockwell and Peters, 1999) and maximum entropy (Maxent; [Phillips et … Related articles – All 2 versions

Exploring the SAWA corpus: collection and deployment of a parallel corpus English-Swahili [PDF] from G De Pauw, PW Wagacha… – Language resources and …, 2011 – Springer … 3). A Maxent classifier (Le 2004) is subsequently trained on this data, … 3 Instances for Maximum Entropy Sentence Aligner … To construct an n-gram language model with the SRILM toolkit (Stolcke 2002), we used the twenty-million-word TshwaneDJe Kiswahili Internet Corpus (de … Cited by 2 – Related articles – All 11 versions

[PDF] The Atlas of Living Australia? s Spatial Portal [PDF] from L Belbin – … 4). Page 5. Figure 4. MaxEnt spatial prediction of Sarcophilus harrisii (Tasmanian Devil) using Aridity index – month max, Precipitation – seasonality and Temperature – annual max mean. … A maximum entropy approach to species distribution modeling. … Related articles – View as HTML

Utilizing gestures to improve sentence boundary detection [PDF] from L Chen… – Multimedia Tools and Applications, 2011 – Springer … data sets [31]. In addition, discriminative modeling approaches, such as maximum entropy (ME) and conditional random field (CRF), were investigated to address weaknesses of the generative HMM approach [32]. As an important … Related articles – All 8 versions

[PDF] Judging grammaticality with tree substitution grammar derivations [PDF] from M Post – Proceedings of the 49th Annual Meeting of the …, 2011 – … Eugene Charniak and Mark Johnson. 2005. Coarse-to- fine n-best parsing and MaxEnt discriminative rerank- ing. In Proc. … Eugene Charniak. 2000. A maximum-entropy-inspired parser. In Proc. … Andreas Stolcke. 2002. SRILM – an extensible language modeling toolkit. In Proc. … Related articles – View as HTML – All 8 versions

Enriching machine-mediated speech-to-speech translation using contextual information VK Rangarajan Sridhar, S Bangalore… – Computer Speech & …, 2011 – Elsevier … View Within Article. 4. Maximum entropy classification. We use a maximum entropy sequence tagging model for the purpose of automatic tagging. … We used the machine learning toolkit LLAMA (Haffner, 2006) to estimate the conditional distribution using Maxent. … Related articles

[PDF] Joint Alignment and Artificial Data Generation: An Empirical Study of Pivot-based Machine Transliteration [PDF] from M Zhang, X Duan, M Liu, Y Xia… – … unlike other feature-based methods, such as CRFs (Lafferty et al., 2001), MaxEnt (Berger et al … based and the proposed synthetic- based strategy are transliteration model- independent while model-based strategy … A Maximum Entropy Approach to Natural Language Processing. … Related articles – View as HTML – All 3 versions

[CITATION] Combining Information Sources for Confidence Estimation with CRF Models MS Seigel… – 2011 Related articles

[PDF] Learning from Chinese-English Parallel Data for Chinese Tense Prediction [PDF] from F Liu, F Liu… – … We chose the maximum entropy (ME) model as the classifier for tense prediction, because ME can … Coarse-to-fine n-best parsing and maxent discriminative reranking. … A word-based approach for modeling and discovering temporal rela- tions embedded in Chinese sentences. … Related articles – View as HTML – All 3 versions

[PDF] MWU-aware Part-of-Speech Tagging with a CRF model and lexical resources [PDF] from M Constant… – ACL HLT 2011, 2011 – … 10CRF++ is an open-source toolkit to train and test CRF mod- els (http://crfpp. … Tagger Model ACC TnT HMM 96.3 TreeTagger Decision trees 96.4 SVMTool SVM 97.2 CRF-STD CRF 97.4 MElt MaxEnt 97.6 CRF-STD+ CRF … A maximum entropy model for part-of-speech tagging. … Cited by 1 – Related articles – View as HTML – All 14 versions

Robust argumentative zoning for sensemaking in scholarly documents [PDF] from S Teufel… – Advanced Language Technologies for Digital …, 2011 – Springer … Finally we feed the feature vectors to the maximum entropy software2) to gen- erate models in training … splitting of sentences) as well as realistic, noisy input (using automatic 2 We use Le Zhang’s toolkit, available at: toolkit.html … Related articles – All 6 versions

[PDF] Expert-Driven Topical Classification of Short Message Streams [PDF] from KY Kamath… – Politics – … MaxEnt is based on the maximum entropy principle [3] and has been widely used for text … principle states to pick the the distribution p? ? C that yields the maximum entropy H(x … Moving to the following day, the classification models could be updated with the sliding window (now … Related articles – View as HTML

Entropic Fragment-Based Approach to Aptamer Design CY Tseng, M Ashrafuzzaman… – Chemical Biology & …, 2011 – Wiley Online Library … We have, however, developed a model system to assess binding to PS using liposome technology and can therefore … a tool for updating information, rather than the MaxEnt will be used. When the new … locations and orientations corresponds to the global entropy maximum and is … Related articles – All 3 versions

[PDF] Extracting Targets and Attributes of Medical Findings from Radiology Reports for Cross-Corpus Search [PDF] from HS Oh, JB Kim… – 2011 – … Table 5. Performance comparison on different sentence offset values (F1 score only) Offset Label SVM NBC MaxEnt … 1996. A maximum entropy approach to natural language processing. Computational linguistics. … A comparison of event models for naive bayes text classification. … Related articles – View as HTML

A Comparison of Different Machine Transliteration Models [PDF] from K Choi, H Isahara… – Arxiv preprint arXiv:1110.1391, 2011 – … In this section, we describe our implementation of the four machine transliteration models (?G, ?P , ?H, and ?C) using three machine learning algorithms: memory-based learning, decision-tree learning, and the maximum entropy model. … 3.3.1 Maximum entropy model … Related articles – All 3 versions

[PDF] Extracting Targets and Attributes of Medical Findings from Radiology Reports in a mixture of Languages [PDF] from HS Oh, JB Kim… – 2011 – … Table 5. Performance comparison on different sentence offset values (F1 score only) Offset Label SVM NBC MaxEnt … 1996. A maximum entropy approach to natural language processing. Computational linguistics. … A comparison of event models for naive bayes text classification. … Related articles – View as HTML – All 2 versions

Beyond Flickr: Not All Image Tagging Is Created Equal JL Klavans, R Guerra, R LaPlante, R Stein… – Workshops at the Twenty- …, 2011 – … The Morphy lemma- tizer from Natural Language Toolkit (NLTK) (Bird, Klein, and Loper 2009 … The second advantage is that a Maximum Entropy model works by ex- tracting features from the … For the Stanford MaxEnt tagger we utilized the bidirectional-distsim-wsj-0-18.tagger … Related articles – All 2 versions

[PDF] Twitter Polarity Classification with Label Propagation over Lexical Links and the Follower Graph [PDF] from M Speriosu, N Sudan, S Upadhyay… – EMNLP 2011, 2011 – … datasets show that our label propagation approach ri- vals a model supervised with … 3.2 Maximum entropy classifier (MAXENT) The OpenNLP Maximum Entropy package7 is used to train … EMOTICON or HCR-TRAIN, henceforth referred to as EMO- MAXENT and GOLDMAXENT … Cited by 1 – Related articles – View as HTML – All 4 versions

Tradeoffs in Accuracy and Efficiency in Supervised Learning Methods [PDF] from L Collingwood… – 2011 – … yi(wT f(xi) + b) = 1 – ?i, ?i = 0. Maximum entropy is an exponential model with a simple intuition: assume nothing about unknown attributes of the data but model all the known attributes. … The maximum entropy model has the following exponential form: p(x) = exp (?T f(x) Z? ) … Related articles

Prioritizing Literature Search Results Using a Training Set of Classified Documents S Matos… – … International Conference on Practical Applications of …, 2011 – Springer … The authors propose a classification and ranking model to evaluate the entire MEDLINE database with respect to a topic … MaxEnt Text+Lemmas+bigrams 0,629 0,655 0,567 0,365 0,491 The Naïve Bayes and Maximum Entropy classifiers where trained using the Mallet toolkit [20 … Related articles – All 2 versions

A comparative study on feature reduction approaches in Hindi and Bengali named entity recognition SK Saha, P Mitra… – Knowledge-Based Systems, 2011 – Elsevier … Bender et al. [2] used count-based feature reduction in a maximum entropy (MaxEnt) based NER system. They … iteratively. In each iteration, some set of candidates were evaluated, and the best ones were added to the model. In Saha et al. … Related articles

[PDF] Modality Specific Meta Features for Authorship Attribution in Web Forum Posts [PDF] from T Solorio, S Pillay, S Raghavan… – … with lexical information by combining the output with a bag- of-words Maximum Entropy classifier and … language model per author and each model generates a perplexity, or cross entropy, value per … E, is the combination of the PCFG with the bag-of-words MaxEnt model, and an … Related articles – View as HTML – All 3 versions

Interruption Point Detection of Spontaneous Speech Using Inter-Syllable Boundary-Based Prosodic Features CH Wu, WB Liang… – ACM Transactions on Asian Language …, 2011 – … in speech. Several methods, including hidden Markov Model (HMM), Maximum Entropy (MaxEnt) and CRF for combining knowledge sources involving both textual information and prosodic cues, were examined. An early work … Related articles

[PDF] Learning Logical Structures of Paragraphs in Legal Articles [PDF] from NX Bach, N Le Minh, TT Oanh… – … the kth model is learned from logical parts in the Layerk of training data, using Conditional random fields (Lafferty et al., 2001; Kudo, CRF toolkit). … problem is how to define the value function f. Our solution is that, first we learn a binary classifier C using maximum entropy model. … Related articles – View as HTML – All 3 versions

[PDF] Utilizing Various Natural Language Processing Techniques for Biomedical Interaction Extraction [PDF] from KM Park, HC Cho… – 2011 – … We implemented the full parsing based argument relation identifier with the maximum entropy model. … For building classifiers, we utilized Zhang le’s MaxEnt toolkit, and the L-BFGS parameter estima- tion algorithm with Gaussian Prior smoothing [22]. 4.1 Experimental Results … Related articles – View as HTML

[PDF] Collaborative ranking: A case study on entity linking [PDF] from Z Chen… – Proc. EMNLP, 2011 – … listwise loss, eg, cross entropy loss(Cao et al., 2007),cosine loss(Qin et al., 2007) … pros: fully consider relationship among objects cons: 1) less well studied in theory selected algorithms Discriminative model for IR (Nallapati, 2004); McRank (Li et al., 2007) … Cited by 1 – Related articles – View as HTML – All 7 versions

Using machine learning for concept extraction on clinical documents from multiple data sources [HTML] from M Torii, K Wagholikar… – Journal of the American …, 2011 – … O},28 indicating each token being the beginning of a concept phrase (B), inside of a concept phrase (I), or outside of a concept phrase (O). Powerful machine learning algorithms for sequence labeling include hidden Markov model (HMM),7 maximum entropy (MaxEnt) tagger,29 … Cited by 1 – Related articles – All 6 versions

Automated classification of psychotherapy note text: implications for quality assessment in PTSD care [PDF] from B Shiner, LW D’Avolio, TN Nguyen… – … of Evaluation in …, 2011 – Wiley Online Library … To determine an appropriate model, ARC auto- matically iterates through various combinations of features and … MaxEnt proved to be slightly better suited for the task than CRFs. … Maximum entropy Canonical token 0.98 0.89 0.93 Canonical token, base token 0.98 0.89 0.93 … Cited by 1 – Related articles – All 3 versions

[PDF] Uncertainty-based Active Learning with Instability Estimation for Text Classification [PDF] from J ZHU… – … 5. EVALUATION 5.1 Settings We utilized a state-of-the-art maximum entropy (MaxEnt) model (Berger et al., … sources into a single, well-grounded statistical model. A publicly available MaxEnt toolkit7 was used in our comparison experiments. To evaluate the … View as HTML

Controlling complexity in part-of-speech induction [PDF] from JV Graça, K Ganchev, L Coheur, F Pereira… – Journal of Artificial …, 2011 – … 5. The idea of replacing the multinomial models of an HMM by maximum entropy models is not new … to exclude word identity features for rare words in order to constrain the model and force … all rare words to the ‘unk’ token in the multinomial setting, the maxent model still captures … Related articles – All 10 versions

[PDF] Tuning as ranking [PDF] from M Hopkins… – Proc. of EMNLP, 2011 – … 1The remainder either did not specify their tuning method (though a number of these used the Moses toolkit (Koehn et al., 2007), which uses MERT for tuning) or, in one … We now detail the three language pairs, two feature scenarios, and two MT models used for our experiments. … Cited by 1 – Related articles – View as HTML – All 7 versions

GeoCAM: A geovisual analytics workspace to contextualize and interpret statements about movement [PDF] from A Jaiswal, S Pezanowski, P Mitra, X Zhang… – Journal of Spatial …, 2011 – … and Maximum Entropy Markov Models (MEMMs) [26], which utilizes the interdepen- dency and compared them with two models that do not, Naive Bayes [22], Maximum Entropy (MaxEnt) [4, 31]. … We trained and evaluated three classifiers, Naive Bayes, Maximum Entropy and … Related articles – All 4 versions

[PDF] Classification methods for finding articles describing protein-protein interactions in PubMed. [PDF] from S Matos… – Journal of integrative bioinformatics, 2011 – … probabilistic and machine learning based classifiers: Naive Bayes (NB), Maximum Entropy (MaxEnt), and Support … used in the texts, it is expected that classification models based on … The Naive Bayes and Maximum Entropy classifiers were implemented and trained using the … Related articles – View as HTML – All 4 versions

[HTML] Quantum correlations in the thermodynamic limit: the XY-model [HTML] from J Batle, A Plastino, AR Plastino, M Casas… – Arxiv preprint arXiv: …, 2011 – … We present a semiclassical model of an atom moving in the evanescent field of a microtoroidal resonator. … Comments: 9 pages. Presented at MaxEnt 2011, The 31st International Workshop on Bayesian Inference and Maximum Entropy Methods in Science and … Related articles – Cached – All 4 versions

Large-scale question classification in cQA by leveraging Wikipedia semantic knowledge L Cai, G Zhou, K Liu… – … of the 20th ACM international conference …, 2011 – … Many researchers have employed machine learning methods (eg, maximum entropy and support vector ma- chine) by using different … Pml is computed using the maximum likelihood es- timator. … for training, and the GIZA++5 toolkit is used to learn the IBM translation model 1. The …

[PDF] Gender attribution: tracing stylometric evidence beyond topic and genre [PDF] from R Sarawgi, K Gajulapalli… – Proceedings of the Fifteenth …, 2011 – … Note that there has been no previous research that directly compares the performance of character- level language models to that of PCFG based models for author attribution, not to mention for gender at- tribution. 4.4 Bag of Words using Maximum Entropy (MaxEnt) Classifier … Cited by 1 – Related articles – View as HTML – All 11 versions

The sum of its parts: reducing sparsity in click estimation with query segments [PDF] from D Hillard, E Manavoglu, H Raghavan… – Information …, 2011 – Springer … We chose a Maximum Entropy (maxent) model to estimate the CTR’s because the learning approach can be efficiently … Maximum likelihood estimation for maxent is a well studied problem (Minka 2003). … for each cluster, and use a 1-in-k encoding as input in the maxent model. … Related articles – All 7 versions

[PDF] Improving Phrase Extraction via MBR Phrase Scoring and Pruning [PDF] from N Duan, M Li, M Zhou… – … Two MaxEnt-based POS taggers (Ratnaparkhi, 1996) are used to tag Chinese and English words … Xiong et al., 2006) with a lexicalized reordering component based on maximum entropy is used … Both the phrase table and the reordering model are trained on the same bilingual … Related articles – View as HTML

[PDF] Segmentation of Broadcast News Story Using Conditional Random Fields and Multimodal Features [PDF] from W Xiaoxuan, XIE Lei, LU Mimi, MA Bin… – … Recently, support vector machine (SVM) [16] and maximum entropy (ME) models [9] have also been used for story segmenta- tion. Despite years of study, most of the previous research has focused on modeling features independently. … View as HTML

[PDF] LEARNING SUB-WORD UNITS AND EXPLOITING CONTEXTUAL INFORMATION FOR OPEN VOCABULARY SPEECH RECOGNITION [PDF] from MC Parada – 2011 – … 78 4.2 From Maximum Entropy to Conditional Random Fields … Arisoy, Can, Parlak, Sak, and Saraclar (2009), which presents a hybrid system for morphologically rich languages with very large vocabulary sizes, concluding that hybrid approaches are capable of modeling a larger … Related articles – View as HTML

[PDF] Focusing on Novelty: A Crawling Strategy to Build Diverse Language Models [PDF] from L Barbosa… – 2011 – … are widely used in text classification: 2 Naive Bayes, Suppport Vector Ma- chines (SVM) and Maximum Entropy (MaxEnt). … SVM per- formed slightly better than MaxEnt for F-measure. … The trigram model is built using a language modeling toolkit [12], with Katz back-off scheme for … Related articles – View as HTML – All 5 versions

Query-focused multi-document summarization: automatic data annotations and supervised learning approaches Y CHALI… – Natural Language Engineering – Cambridge Univ Press … the MaxEnt technique to text summarization and found that the maximum entropy classifier produces … supervised machine learning techniques, viz., SVMs, HMMs, CRFs, and MaxEnt, and two … ROUGE6 toolkit (Lin 2004) to measure sentence goodness against model summaries. … Related articles – All 2 versions

Improving Korean verb-verb morphological disambiguation using lexical knowledge from unambiguous unlabeled data and selective web counts S Kim, J Yoon, J Seo… – Pattern Recognition Letters, 2011 – Elsevier … 326 = = == ≥= >â‹…<−= N i iii jij N i N j iji N i iD yts yyL 1 1 11 0,0.. 2 1 αα ααα xx (2) 327 For SVM learning, the LibSVM toolkit was used after construction of a feature vector … For ME learning, we use the Maxent toolkit9. … A maximum entropy model for part-of-speech tagging …

[PDF] CUNY-BLENDER TAC-KBP2010 [PDF] from Z Chen, S Tamang, A Lee, X Li, WP Lin… – … We use the Alchemy toolkit (Kok et al., 2007) to encode inference rules such as those based on traversing family trees. … We develop a Maximum Entropy (MaxEnt) based supervised re-ranking model to re-rank candidate answers for the same slot. … Related articles – View as HTML – All 4 versions

The IBM speech-to-speech translation system for smartphone: Improvements for resource-constrained tasks B Zhou, X Cui, S Huang, M Cmejrek… – Computer Speech & …, 2011 – Elsevier … 4 (Brown et al., 1993), and the approaches based on hidden Markov models (HMM) (Vogel … GIZA++ (Och and Ney, 2000) is a publicly available toolkit to extract word alignments from … In practice, we normally enforce a maximum phrase length during phrase extraction to reduce … Related articles

Domain adaptation for text categorization by feature labeling [PDF] from C Kadar… – Advances in Information Retrieval, 2011 – Springer … L-BFGS, a quasi-Newton optimization method that estimates the model parameters. … Dataset # source labeled instances MaxEnt # source labeled features TransferLF on source TransferLF … from the source domain data only: a standard super- vised maximum entropy classifier as … Related articles – All 4 versions

Knowledge-based nonuniform sampling in multidimensional NMR AD Schuyler, MW Maciejewski, H Arthanari… – Journal of biomolecular …, 2011 – Springer … The entropy component of Eq. … Rather than assigning a peak to the location of the local maximum value (as done presently), it is possible to compute an off-grid … Second, the nonlinear scaling introduced by MaxEnt may additionally skew overlapping peak locations (Kubat et al. … Related articles – All 4 versions

[PDF] An interdisciplinary VR-architecture for 3D chatting with non-verbal communication [PDF] from S Gobron, J Ahn, Q Silvestre, D Thalmann, S Rank… – … Specifically, we use three standard, probabilistic, state-of-the-art classifiers: a Language Model [PSW03], a Naive Bayes [MS99] and a Maximum Entropy [NLM99] classifier. … document D the outcome of the classifiers is {Px(ob j|D),Px(pos|D)} where x = {LM,NB,MaxEnt} for each … Related articles – View as HTML – All 3 versions

Machine-learned solutions for three stages of clinical information extraction: the state of the art at i2b2 2010 [HTML] from B de Bruijn, C Cherry, S Kiritchenko… – Journal of the American …, 2011 – … Collins M. . Discriminative Training Methods for Hidden Markov Models: Theory and Experiments with Perceptron Algorithms. … [Web of Science]. ?: Ratnaparkhi A. . A Maximum Entropy Part-of-Speech Tagger. … (accessed 21 Jun 2010). … Cited by 2 – Related articles – All 8 versions

[PDF] Learning to transform and select elementary trees for improved syntax-based machine translations [PDF] from B Zhao, YS Lee, X Luo… – Proc. of ACL, 2011 – … SCFG in general; in section 3, our proposed approach is explained in detail, including the statistical operators using a MaxEnt model; in section … for the given source tree ? together with its neighborhood ¯m. Here, in our approach, we applied a Maximum Entropy model, which is … Cited by 2 – Related articles – View as HTML – All 10 versions

Progressive Alignment and Discriminative Error Correction for Multiple OCR Engines [PDF] from WB Lund, DD Walker… – Document Analysis and …, 2011 – … for training a maximum entropy model to select correct hypotheses from the aligned OCR sequences. …, 2008. [3] H. Baird. The State of the Art of Document Image Degradation Modelling. … Coarse-to-fine n-best parsing and MaxEnt discriminative reranking. … Cited by 1 – Related articles – All 3 versions

[PDF] Single and multi-objective optimization for feature selection in anaphora resolution [PDF] from S Saha, M Poesio, A Ekbal… – 2011 – … ranking), has interfaces to different machine learners (MaxEnt, SVM, deci- sion trees) and provides a … In the mention-pair model, anaphora resolution is recast as a binary classifica- tion problem … We train a maximum entropy classifier and fol- low the approach of (Soon et al., 2001 … Cited by 1 – Related articles – View as HTML – All 3 versions

[PDF] Appraisal Expression Recognition with Syntactic Path for Sentence Sentiment Classification [PDF] from Y Zhao, B Qin, W Che… – World, 2011 – … A Maximum Entropy (ME) [28] Modeling Toolkitf is used for the common feature-based method, our appraisal expression based (AppExp-based) method and the Combination method. … f toolkit.html Page 13. … Cited by 1 – Related articles – View as HTML – All 8 versions

Named Entity Recognition F Béchet – Spoken Language Understanding, 2011 – Wiley Online Library … However, like for any semantic model, there is no consensus about what should be the ideal set of categories to use in a NER task. … The two ways of dealing with them is either to remove them prior to NER or to modify the NER models in order to integrate such disfluencies. … Cited by 1 – Related articles

[PDF] Multi-aspect Sentiment Analysis with Topic Models [PDF] from B Lu, M Ott, C Cardie… – … [8] present MaxEnt-LDA, a maximum entropy hybrid model that discovers both aspects and aspect-specific opinion words. … 1In this work we sample all models for 1,000 iterations, with a 500- iteration burn-in and a sampling-lag of 10. … View as HTML

[PDF] Multi-metric optimization for coreference: The UniTN/IITP/Essex submission to the 2011 CONLL Shared Task [PDF] from O Uryupina, S Saha, A Ekbal… – ACL HLT 2011, 2011 – … It also has interfaces to different machine learners (MaxEnt, SVM, de- cision trees). … Again be- cause of time considerations, we used decision trees as implemented in Weka as our classification model instead of maximum-entropy or SVMs. … Related articles – View as HTML – All 10 versions

Computational Aspects of Lexical Functional Grammar M Forst – Language and Linguistics Compass, 2011 – Wiley Online Library … Since the late 1990s, finally, machine-learned models (typically maximum entropy models, support vector machines or multi-layered artificial neural networks) based on basically arbitrary learning that can be extracted from the syntactic analyses features (also called ‘properties … Related articles

Predicting speculation: A simple disambiguation approach to hedge detection in biomedical literature [HTML] from E Velldal – Journal of Biomedical Semantics, 2011 – … [3], where a maximum entropy (MaxEnt) classifier is … [3], but is set within the framework of Support Vector Machine (SVM) classification [8] instead of MaxEnt. … test for paired samples, which in our setting considers how often the classifier decisions of two given models differ (with … Cited by 2 – Related articles – Cached

Sensitivity enhancement for maximally resolved two-dimensional NMR by nonuniform sampling D Rovnyak, M Sarcone… – Magnetic Resonance in …, 2011 – Wiley Online Library … so that NUS in two dimensions presents no added difficulty in the performance of MaxEnt. … be improved by about 5% if the NUS was a better model of the … in the frequency domain obtained by spectral estimation techniques such as maximum entropy reconstruction, although the … Cited by 1 – Related articles – All 2 versions

Insulin-like growth factor binding protein-2: NMR analysis and structural characterization of the N-terminal domain CA Galea, M Mobli, KA McNeil, TD Mulhern… – Biochimie, 2011 – Elsevier … NUS datasets were processed using the maximum entropy (MaxEnt) reconstruction algorithm implemented in the Rowland NMR Toolkit processing environment, using automatically determined … 2.5. Secondary structure prediction and 3D structure homology modelling. … Related articles – All 3 versions

Integrating statistical genetic and geospatial methods brings new power to phylogeography [PDF] from LM Chan, JL Brown… – Molecular Phylogenetics and Evolution, 2011 – Elsevier … the WorldClim database 6 (Carnaval and Moritz, 2008; Hijmans et al., 2005) in Maxent 3.3.2 … Robertson et al., 2009; Figure 3A)were 9 used to generate species distribution models for four time periods – present day, 10 mid-Holocene (6 kya), last glacial maximum (LGM, 21 … Cited by 4 – Related articles – All 9 versions

[PDF] A stochastic collocation method for uncertainty quantification and propagation in cardiovascular simulations [PDF] from S Sankaran… – Journal of Biomechanical …, 2011 – … of input PDFs can be made using data-driven techniques such as MaxEnt [25 … with roughly 80,000 ele- ments, which ensured spatial convergence, was generated for each AAA model. … For more complex input data, maximum entropy techniques can be used to compute the input … Cited by 4 – Related articles – View as HTML – All 4 versions

[PDF] Integrating Source-Language Context into Log-Linear Models of Statistical Machine Translation [PDF] from R Haque – 2011 – … TM Translation Model LM Language Model MERT Minimum Error-Rate Training HMM Hidden Markov Model AER Alignment Error Rate SVM Support Vector Machine MaxEnt Maximum Entropy LDC Linguistic Data Consortium BTEC Basic Travel Expression Corpus … Related articles – View as HTML – All 3 versions

[PDF] Evaluating the impact of coder errors on active learning [PDF] from I Rehbein… – Proceedings of the 49th Annual Meeting …, 2011 – … 3 4The MaltParser: 5The split has been made … 0.735 0.718 entropy – ALbias 0.806 0.793 0.759 0.748 0.702 0.651 0.625 0.630 0.622 entropy y ALbias … In the next step we train n = 5 maximum en- tropy classifiers on subsets of … Related articles – View as HTML – All 9 versions

Developing methodology for Korean particle error detection [PDF] from M Dickinson, R Israel… – Proceedings of the 6th Workshop on …, 2011 – … We use the Maximum Entropy Toolkit (Le, 2004) for machine learning … Table 1 shows the results of the MaxEnt system for step 1, using training data built for the topics in … the ones implemented in Criterion (see Leacock et al., 2010), as well as a language model approach (Gamon … Related articles – All 13 versions

Landscape-scale dynamics and restoration of dryland forest ecosystems A Newton, E Cantarello, N Tejedor… – … and Practice of …, 2011 – … II ‘! El-Ii 55-III The above maps illustrate LANDIS-11 model output, for the different disturbance scenarios con- ducted in QIIIIPUC, Chile, over a period of ZOO years. The different colours on the maps refer to different maximum ages of the forest stands. … Related articles – All 3 versions

Regulation of alternative splicing by the core spliceosomal machinery [HTML] from AL Saltzman, Q Pan… – Genes & Development, 2011 – … for exons profiled by RNA-seq were calculated using maximum entropy models (Yeo and … Massachusetts Institute of Technology ( Xmaxentscan_scoreseq.html … criteria (n = 3173) using Web-based Gene Set Analysis Toolkit ( … Cited by 7 – Related articles – All 5 versions

[PDF] A Memory-Based Tagger for Polish [PDF] from TS Adam Radziszewski – … that some improvement could be obtained by removing less significant attributes from the model. … (Sniatowski, 2011) adds a fourth tagger, a pro- totype maximum entropy tagger of … RFTagger 89.78% 0.05% 3-way Voting 93.96% 0.06% WMBT 93.16% 0.09% maxent 89.70% 0.12 … Related articles – View as HTML

diviText: visualizing text segmentation for text mining. [PDF] from AC Jones – 2011 – … Maximum Entropy w ximum intropy @w xintA l ssifiers use st tisti l methE ods of multinomi l logisti regression to orre t for the unre listi … toolkit for n tur ll ngu ge pro essing through st tisti sF e few p ilities in lude text lusteringD do ument l ssifi tionD nd topi modeling @wg llumD …

A multiobjective simulated annealing approach for classifier ensemble: Named entity recognition in Indian languages as case studies A Ekbal… – Expert Systems with Applications, 2011 – Elsevier … Diverse classification methods such as Maximum Entropy (ME), Conditional Random Field (CRF) and Support Vector Machine (SVM) are used to build different models depending upon the various representations of the available features. … Related articles

Affirmative Cue Words in Task-Oriented Dialogue A Gravano, J Hirschberg… – Computational Linguistics, 2011 – MIT Press … subclass of cue phrases or dialogue acts. Some have attempted to model other types … diately following w (see Table 4). POS tags were labeled automatically for the whole corpus using Ratnaparkhi’s (1996) maxent tagger trained on a subset of the Switchboard … Cited by 2 – Related articles – All 2 versions

Natural Language Processing to the Rescue? Extracting “Situational Awareness” Tweets During Mass Emergency [PDF] from S Verma, S Vieweg, WJ Corvey, L Palen… – Fifth International AAAI …, 2011 – … Therefore, automatic topic modeling of tweet content is an ongoing interest (Ramage et al., 2010). 386 Page 3. … Classification Methods We experimented with two standard machine-learning methods for classification, Naïve Bayes (NB) and Maximum Entropy (MaxEnt). … Related articles – All 3 versions

[PDF] NewsX [PDF] from M Wunderwald… – 2011 – … The typical representation of a stochastic model is a probabilistic network that represents … statistical evidence from different sources, that is based on the Maximum Entropy (ME) Principle. … call multinominal logistic regressions (in fields of language processing also called MaxEnt). … Related articles – View as HTML

Multiobjective Simulated Annealing Based Approach for Feature Selection in Anaphora Resolution A Ekbal, S Saha, O Uryupina… – Anaphora Processing and …, 2011 – Springer … ranking), has interfaces to different machine learn-ers (MaxEnt, SVM, decision trees) and provides a … pair model without rank- ing as in [19].In the mention-pair model, anaphora resolution … We train a maximum entropy classifier and follow the approach of [19] to partition mentions …

Multiobjective optimization for classifier ensemble and feature selection: an application to named entity recognition A Ekbal… – International Journal on Document Analysis and …, 2011 – Springer … The probability distribution that satisfies the above property is the one with the highest entropy. It is unique, agrees with the maximum likelihood distribution, and has the exponential form … We use the Generalized Iterative Scaling [45] algorithm to estimate the MaxEnt parameters. … Cited by 1 – Related articles

[PDF] A De-identifier For Electronic Medical Records Based On A Heterogeneous Feature Set [PDF] from A Tafvizi – 2011 – … 46 8 Page 9. 4.3 Model Construction . . . . . 46 … 60 7.2.2 More Sophisticated Text Pre-Processing . . . . . 60 7.2.3 The Two-Stage Model . . . . . 60 7.2.4 Feature Selection . . . . . … Related articles – View as HTML

Parallel sentence generation from comparable corpora for improved SMT S Abdul Rauf… – Machine Translation, 2011 – Springer … Furthermore, instead of a max- imum entropy (MaxEnt) classifier, simple measures such as word error … They use sentence- and lexicon-based methods com- bined under a maximum likelihood criterion. … a sentence pair is parallel or not is performed by a MaxEnt classifier trained … Cited by 1 – Related articles

[DOC] Conservation and Sustainable Use of PGRFA: a Toolkit for National Strategy Development [DOC] from N Maxted, JM Brehm… – … HYPERLINK “D:Pos_PhD2010_2011_FAO_Toolkittoolkittoolkittoolkit_August2011 PGRFA_Toolkit_DRAFT_28_08_11_joana3.doc” l “_Toc302507174” **Box 15. SPECIES DISTRIBUTION MODELS * PAGEREF _Toc302507174 h **98**. … View as HTML

[PDF] Exploiting Parse Structures for Native Language Identification [PDF] from SMJ Wong… – … For our experiments we used a maximum entropy (MaxEnt) machine learner, MegaM4 (fifth release) by Hal Daumé III. … more quickly as number of features in- creased, so we only report the MaxEnt results here). The classifier is tuned to obtain an optimal classifi- cation model. … Cited by 1 – Related articles – View as HTML – All 6 versions

[PDF] SEARCHING A WEBSITE VIA SPOKEN QUESTIONS [PDF] from M Durukal – 2011 – … HMI Human Machine Interface HMM Hidden Markov Model HTML Hyper Text Markup Language IDF Inverse Document Frequency IR Information Retrieval LG Language Generation MALLET MAchine Learning for LanguagE Toolkit QA Question Answering … Related articles – View as HTML

A statistical confidence measure framework for probabilistic parsing [PDF] from R SÁNCHEZ SÁEZ – 2011 – … In that work, confidence is measured by Rényi entropy. … Thus, the parse tree is obtained through a chosen parsing algorithm which uses a stochastic model. … used in parsing for sev- eral tasks, for example, learning the stochastic information of PCFGs by maximum log-likelihood. …

[PDF] Joint bilingual sentiment classification with unlabeled parallel corpora [PDF] from B Lu, C Tan, C Cardie… – … of the 49th Annual Meeting of …, 2011 – … Maximum entropy (MaxEnt) models1 have been widely used in many NLP tasks (Berger et al., 1996; Ratnaparkhi, 1997 … Maximum likelihood parameter estimation (training) for such a model, with a set of labeled examples , amounts to … With MaxEnt, we learn from the input data: … Cited by 2 – Related articles – View as HTML – All 11 versions

Rotating Stars and Revolving Planets: Bayesian Exploration of the Pulsating Sky [PDF] from TJ Loredo – Arxiv preprint arXiv:1107.5805, 2011 – … Set the test size a (the maximum p-value we will accept as indicating the actual signal is not … Physicists know its logarithm as the configura- tional entropy of the {nj }. In fact, I devised this model specifically to obtain this result, formalizing a clever intuition of Gregory’s that entropy … Cited by 1 – Related articles – All 5 versions

[PDF] Toward Improving the Automated Classification of Metonymy in Text Corpora [PDF] from FMO Ferraro – 2011 – … Using a pre-existing dataset of thousands of samples, we formulate the hypothesis that a better modeling of the underlying syntactic, semantic and conceptual meanings within a document (sample) can aid automated metonymy classification. … Related articles – View as HTML – All 3 versions

[PDF] Semantic Role Labeling Without Treebanks? [PDF] from SA Boxwell, C Brew, J Baldridge, D Mehay… – … The system is organized in a two- stage pipeline of maximum entropy models3, fol- lowing the … ARG1, and so on (figure 4). The identification model and the classification model share the … 3We use the Zhang Le maxent toolkit, available at … Related articles – View as HTML – All 3 versions

Identifying relevant data for a biological database: Handcrafted rules versus machine learning [PDF] from AK Sehgal, S Das, K Noto… – … IEEE/ACM Transactions …, 2011 – … We use a maximum-entropy classifier, implemented as part of the MALLET toolkit [27].6 For the … 5. We also evaluated the naive Bayes classifier, using the Rainbow toolkit [26]. … 6. MALLET also has an implementation of naive Bayes but the discriminative MaxENT model proved to … Cited by 5 – Related articles – All 9 versions

Named entity recognition in Bengali and Hindi using support vector machine A Ekbal… – Lingvisticae Investigationes, 2011 – … 1999)), Maximum Entropy (ME)(New York University’s MENE (Borthwick 1999; Borthwick et al. … 2000) introduced a hybrid system by combining HMM, MaxEnt and handcrafted … The SVM model has an equivalent dual formulation, characterized by a weight vector a and a bias b. In … Related articles – All 2 versions

Rotating Stars and Revolving Planets: Bayesian Exploration of the Pulsating Sky JM Bernardo, MJ Bayarri, JO Berger… – Bayesian …, 2011 – … Set the test size a (the maximum p-value we will accept as indicating the actual signal is not … Physicists know its logarithm as the configura- tional entropy of the {nj}. In fact, I devised this model specifically to obtain this result, formalizing a clever intuition of Gregory’s that entropy …

Filtering and clustering relations for unsupervised information extraction in open domain [PDF] from W Wang, R Besançon, O Ferret… – Proceedings of the 20th ACM …, 2011 – … Classically, we trained a Naive Bayes classifier, a Maximum Entropy classifier (MaxEnt), a Decision Tree and a Support Vector Machine classifier (SVM). The first three models were implemented using the tools provided by MALLET [22] while the last model was implemented …

[PDF] Inspecting Sample Reusability for Active Learning [PDF] from K Tomanek… – … For AL with a NB and a MaxEnt-based selector, the confidence is estimated as the margin between … For maximum margin classification, the decision value d(x) = <w,x> + b indicates the distance of an … Entropy utility function quantifies disagreement (Engelson and Dagan, 1996) … Cited by 1 – Related articles – View as HTML – All 2 versions

[PDF] Email formality in the workplace: a case study on the Enron corpus [PDF] from K Peterson, M Hohensee… – ACL HLT 2011, 2011 – … 3 Performance of the formality classifier We trained a Maximum Entropy (MaxEnt) classi- fier in the Mallet package … 2 Performance of the request classifier We trained a MaxEnt classifier and ran 10-fold cross … Constructing Business Email Messages: A Model of Writer’s Choice. … Related articles – View as HTML – All 13 versions

[PDF] Goodness: A method for measuring machine translation confidence [PDF] from N Bach, F Huang… – … Annual Meeting of the Association for …, 2011 – … The parallel sentences have word alignment au- tomatically generated with HMM and MaxEnt word aligner (Ge, 2004; Ittycheriah and Roukos, 2005). Bilingual phrase translations are extracted from these word-aligned parallel corpora. The language model is a 5-gram model … Cited by 1 – Related articles – View as HTML – All 10 versions

[PDF] Basic Design of the architecture and methodologies (second round) [PDF] from G Rigau, B Magnini… – … 3.4.4 English • XML input/output added to RASP system. • Multiword recognizer: data extracted and integrated. • NE Recognizer: maximum entropy model-based, trained on CONNL data, and inte- grated. 3.4.5 Italian • NE Recognizer: porting to Italian done. … Related articles – View as HTML – All 4 versions

Turn-taking cues in task-oriented dialogue [PDF] from A Gravano… – Computer Speech & Language, 2011 – Elsevier … We examine the same question for backchannels. In Section 2 we discuss previous work on modeling turn-taking and on end-of-turn detection. … As part of a larger project on modeling discourse structure in American English, Jurafsky et al. … Cited by 10 – Related articles – All 8 versions

[PDF] Effective measures of domain similarity for parsing [PDF] from B Plank… – Proceedings of the 49th Annual Meeting of the …, 2011 – … a Language Model based on a scoring mechanism that compares cross- entropy scores. … is a property that “makes it unsuitable for distributions derived via maximum- likelihood estimates … proach on English PoS tagging using two different taggers: MXPOST, the MaxEnt tagger of … Cited by 2 – Related articles – View as HTML – All 14 versions

[PDF] Identifying the Gist of Conversational Text: Automatic Keyword Extraction and Summarization [PDF] from F Liu, Y Liu, C Busso, S Harabagiu… – 2011 – … performance. These properties were incorporated in both unsupervised Maximum Marginal … automatically collected a large set of noisy training pairs using a novel web-based approach, and aligned them at the character level for modeling training. For Twitter topic summariza- … Related articles – View as HTML – Library Search – All 3 versions

The Yale cTAKES extensions for document classification: architecture and application [HTML] from V Garla, VL Re, Z Dorey-Stein… – Journal of the …, 2011 – … Extraction System (caTIES).10-13 The open-source WEKA data-mining toolkit has been … cTAKES uses the OpenNLP Maximum Entropy package for sentence detection, tokenizing, part-of-speech tagging … the online appendix for a detailed description of the data model.28 YTEX … Related articles – All 7 versions

Molecular evolution and historical biogeography of new world birds [PDF] from BT Smith – 2011 – … variables (BIO1, BIO2, BIO5, BIO6, BIO8, BIO9, BIO10, BIO12, BIO13, BIO15 and BIO18). I generated five replicate models for each separate clade identified from our phylogenetic analysis using the maximum entropy algorithm in MAXENT 3.3.3 (Phillips et al. 2006). … Related articles – All 2 versions

Information extraction from chemical patents [PDF] from DM Jessop – 2011 – … JVM Java Virtual Machine MEMM Maximum Entropy Markov Model NLP Natural Language Processing … researcher’s lab book, effectively lost to the community, while the traditional business model of a journal requires the erection of paywalls. … toolkits in the Unilever Centre. … Cited by 1 – Related articles

A parser-based approach to detecting modification of biomedical events A MacKinlay, D Martinez… – … workshop on Data and text mining …, 2011 – … classifiers described in Section 2 as a source of trigger annotations, selecting the combination with the best performance over the development set. We used a maximum entropy classifier, by applying Zhang Le’s Maxent Toolkit.4 … Cited by 1

[PDF] Unterstützung kollaborativer Schreibprozesse durch ausgewählte Verfahren der Computerlinguistik [PDF] from S Groß, N Pinkwart… – … II Page 6. API Application Programming Interface, Programmierschnittstelle CSCW Computer Supported Cooperative Work, computerunterstützte Gruppenarbeit ERM Entity-Relationship- Model, Gegenstand-Beziehung-Modell FIFO First In – First Out … Related articles – View as HTML – All 2 versions

[PDF] Estimation d’un modèle de traduction à partir d’alignements mot-à-mot non-déterministes [PDF] from NTAAF Yvon – … Keywords: statistical machine translation, phrase based translation models, word alignment models. … PostCAT (Posterior Constrained Alignment Toolkit) propose une implémentation des modèles HMM permettant d’injecter … à la section 3. Il s’agit d’un classifieur MaxEnt qui prédit … Related articles – View as HTML – All 8 versions

[PDF] Personal sense in subjective language research in the blogosphere [PDF] from P Panicheva – … judgments about quality. The model was trained with the Sequential Minimal Optimization [Platt, 1998] algorithm. … personality traits, using n-grams as features, and proceed to test the resulting model on a larger corpus in [Oberlander et al, 2007]. … View as HTML

[PDF] Recognition of Multiple Speech Sources using ICA [PDF] from E Hoffmann, D Kolossa… – … where H(g(Y)) is the joint entropy of g(Y). This is known as the information maximization approach … In [36] the permu- tation problem was solved using a maximum-likelihood-ratio criterion between the adjacent frequency … For this purpose, the following signal model is assumed … Related articles – View as HTML – All 3 versions

Integrating source-language context into phrase-based statistical machine translation R Haque, SK Naskar, A van den Bosch… – Machine Translation, 2011 – Springer … (2007) propose an SMT architecture based on stochastic finite state transducers that addresses global lexical selection in which parameters are discriminatively trained using a MaxEnt model consider- ing n-gram features from the source sentence. Ittycheriah et al. … Related articles