author: alexander gelbukh

... feats Dictionary-based + SVD (U) + 3-feats Supervised 4-scores + 3-feats ... All of the embeddings are learnt on four corpora separately. We have used ... 3.5. As can be seen, at least one of our methods outperforms the baseline word2vec ...

... smallest value of p for which G ( p ) has no positive - distance cycles ... sequence of longest- distance trees such that feasible p values decrease from one tree to another until a positive ... Scheduling Algorithms for Interval-Valued Data.

... real-life games. The data are available for download in text form. In the data from March 26, 2010 there are 745,030 pairs from the Free Associations and 1,199,235 pairs from Categorilla and Categodzilla. In our case the players' task ...

... tn, leila.baccour@edunet.tn, adel.alimi@ieee.org 2 Electrical and Computer Engineering Dept, University of West Florida, 11000 University Parkway, Pensacola, FL 32514, USA mkhabou@uwf.edu Abstract. We present an indexing and retrieval ...

... points to obtain standard SVM from RSVM, which results in ... Tutorial on Support Vector Machines for Pattern Recognition, Data Mining and ... J. S.: An Introduction to Support Simplify Decision Function of Reduced Support Vector Machines 441.

... pronouns to improve the accuracy of their transla- tions to both Spanish and English in an underlying MT project . 1 Introduction Pronoun resolution – the task of ... Portuguese Pronoun Resolution: Resources and Evaluation Introduction.

... excitatory synapse to the Motor through the FWD neuronal population that kept the robotic device moving forward. Each time a neuronal population corresponding to a wheel fired: an excitatory synapse was sent to the Motor, an inhibitory ...

20th Mexican International Conference on Artificial Intelligence, MICAI 2021, Mexico City, Mexico ... Barbaro Ferro Karina Figueroa Denis Filatov Dora-Luz Flores Juan Jose Flores ... Mexico Universidad Michoacana de San Nicolás de Hidalgo ...

... (Musto et al. 2014; Cambria et al. 2012). – SenticNet. SenticNet is resources for English language which is used to ... Engineering Feature selection is divided on different categories: Adaptation of Sentiment Analysis Techniques to ...

... download the tool can be found in [2]. For the LM we used the CMU-Cambridge ... context-free grammar, then it generates sentences by traversing the grammar ... 3-gram models, both showing very similar results. This was not surprising as ...

... words talking, talked, talk and talks to the word, talk. The major task of a ... list of endings while some complex stemmer requires morphological knowledge ... adjectives, adverbs, pronouns and proper names. A suffix list of 18 suffixes ...

... Error-Repair Parsers* Carlos Gómez-Rodríguez1, Miguel A. Alonso1, and Manuel Vilares2 2 1 Departamento de Computación, Universidade da Coruña (Spain) {cgomezr,alonso}@udc.es Escuela Superior de Ingeniería Informática, Universidade de ...

... best describes the issue, i.e., an aspect specific opinion phrase on the head aspect and containing the head aspect. Throughout the paper, we will refer to head aspect and issue subject interchangeably. Examples below show labeled product ...

... Morningstar 733 Camden Property 203 National Instruments 712 Capital One 997 NetApp 800 CarMax 959 Nordstorm 839 ... Glassdoor numbered less than 500. 2 Reviews for each company in the dataset were selected based on them bearing the most ' ...

... emails in German and English. They developed two models: for the first one they translated German emails into English and then applied opinion-bearing words. For the second one they translated English opinion-bearing words into German ...

16th International Conference, CICLing 2015, Cairo, Egypt, April 14-20, 2015 ... Telugu. (c)Agent Based Systems: These systems allow more natural flow of ... calendar task domain in [5]. Our model can be categorized as an agent based ...

... Test 1 86.07 97.37 74.14 79.12 98.00 Test 2 86.07 97.37 74.14 79.12 98.00 Test 3 83.79 97.02 69.71 76.37 97.86 Test ... parts of speech in the text studied. Meaning, how much could one find out about the subjectivity of a text when only ...

... example, cue phrases such as “because”, “due to”, “if ... then”, and “this caused” often mark relations between cause and effect statements. Naturally, there are many other causal relations that are not marked explicitly in textual ...

... kj j j k N kj j k j ( ) ChainConnectivityc , || j c where r(wij, cj) is the number of relations that include wij in chain cj, | cj | is the number of unique words in chain cj, and Į is a constant of small value. WordConnectivity (wij) ...

... cantemos ( let's sing ) — . On the other hand , we will speak of derivational morphemes when they effect a semantic change on the base and , often , also effecting a change of syntactic class —aburrir ( to bore ) , aburrimiento ...

... verbs to be or to have . We want to point out that we are not implying by ... most frequent co - occurrences : the 100 most frequent noun - verb relations ... common elements and the number of elements that differ between two classes of ...

... Tongyici Cilin that is the “HIT-CIR Tongyici Cilin (Extended)”. Since a word might have several meanings, then it might contain multiple semantic items. And word similarity is obtained by choosing the maximum similarity from multiple ...

... time series, kNN performs as good as ARIMA in the onestep-ahead, iterative ... R and JavaScript. Apress (2013) 3. Brookwell, P.J., Davis, R.A. ... Introductory Time Series with R. Springer (2009) 6. Douglas, C., Montgomery, C.L.J. ...

... rules are likely to provide very good performance. However, to obtain the best performance a refined knowledge base ... Backgammon Rules. http://www.bkgm.com/rules.html 9. Key Concepts in Backgammon. http://www.redtopbg.com/keyconcepts ...

... bad synonym is shown in Figure3(b) where a list of proper names are proposed (Katia) rather than the correct candidate (Nawarah). Table4also show different system combination that we attempted. In system A + B +C and A+B+C+Freq5 ...

... frame rates of subtitle pairs do not match. This can however be easily ... encoding. In order to overcome this, we converted all files to UTF-8 ... correctly describes the language. Fig. 2. An example for the available noise in subtitle ...

... George Mason University , Fairfax , Virginia 22030 U.S.A. jbala@gmu.edu Abstract . This paper introduces a multistrategy learning approach to the cate- gorization of text documents . The approach benefits from two existing , and in our ...

... mimix ( 7 ) idee che sprigionano calore ClamAV website has Moved ! ClamAV : Project News ClamAV : Binary packages ... MIMIX.com ~ High Availability , Data Protection , Disaster Recovery , Replication Domain not defined Domain web Domain ...

... copy Synonym Changing structure WD_System 2 WS_System P_System 6 8 2 6 8 1 4 4 The experimental results show that the proposed P_System produces more precise results in exact copy and changing synonym. It shows that in the proposed ...

... neoclas- sical elements. One or more ICFs can appear sequentially along with one Fig. 1. Algorithm for aligning neoclassical compounds. Cs = source. Neoclassical Compound Alignments from Comparable Corpora 75 Handled Forms.

... Declaration of Independence 2/17author of the Declaration Of Independence was educated at William and Mary College key author of the Declaration of Independence principal author of the Declaration of Independence 3/10best educated and ...

... Hindi word: 'aarop'. The target translation could lose its meaning if one does not take the term translation and ... assess theirs MT systems in the matter of the domain-specific term translation, and be a blessing for the translation ...

... or 1), the dataset then becomes 5,511 positive answers and 4,471 negative answers. For example, a question used to test Biology concepts is “Qual a diferença entre a célula animal e a célula vegetal?”, and the correct answer expected ...

... bankP thein bankP ≠ . Akiba [6] proposed using selective backoff. Their approach is similar to ours in that they use backoff with two different models. One of the models is probabilistic model and the other is a grammar network. The ...

... synonyms, superordinates/hyponyms, opposite words, ellipsis, reference words and connectives. These cohesive devices are categorized into four groups: reiterative devices, reference words, ellipsis and cue phrases. The Reiterative ...