• Contenu principal
  • Menu
OpenEdition Books
  • Accueil
  • Catalogue de 16091 livres
  • Éditeurs
  • Auteurs
  • Facebook
  • X
  • Partager
    • Facebook

    • X

    • Accueil
    • Catalogue de 16091 livres
    • Éditeurs
    • Auteurs
  • Ressources numériques en sciences humaines et sociales

    • OpenEdition
  • Nos plateformes

    • OpenEdition Books
    • OpenEdition Journals
    • Hypothèses
    • Calenda
  • Bibliothèques

    • OpenEdition Freemium
  • Suivez-nous

  • Lettre d’information
OpenEdition Search

Redirection vers OpenEdition Search.

À quel endroit ?
  • Accademia University Press
  • ›
  • Collana dell'Associazione Italiana di Li...
  • ›
  • Proceedings of the Fifth Italian Confere...
  • ›
  • Contributed Papers
  • ›
  • Sentences and Documents in Native Langua...
  • Accademia University Press
  • Accademia University Press
    Accademia University Press
    Informations sur la couverture
    Table des matières
    Liens vers le livre
    Informations sur la couverture
    Table des matières
    Formats de lecture

    Plan

    Plan détaillé Texte intégral 1 Introduction 2 The Classifier and Features 3 Experiments and Results 4 Stacked Classifier 5 Conclusions 6 Aknowledgments Bibliographie Notes de bas de page Auteurs

    Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-it 2018

    Ce livre est recensé par

    Précédent Suivant
    Table des matières

    Sentences and Documents in Native Language Identification

    Andrea Cimino, Felice Dell’Orletta, Dominique Brunato et Giulia Venturi

    p. 125-130

    Résumés

    Starting from a wide set of linguistic features, we present the first in depth feature analysis in two different Native Language Identification (NLI) scenarios. We compare the results obtained in a traditional NLI document classification task and in a newly introduced sentence classification task, investigating the different role played by the considered features. Finally, we study the impact of a set of selected features extracted from the sentence classifier in document classification.

    Partendo da un ampio insieme di caratteristiche linguistiche, presentiamo la prima analisi approfondita del ruolo delle caratteristiche linguistiche nel compito di identificazione della lingua nativa (NLI) in due differenti scenari. Confrontiamo i risultati ottenuti nel tradizionale task di NLI ed in un nuovo compito di classificazione di frasi, studiando il ruolo differente che svolgono le caratteristiche considerate. Infine, studiamo l’impatto di un insieme di caratteristiche estratte dal classificatore di frasi nel task di classificazione di documenti.

    Texte intégral 1 Introduction 2 The Classifier and Features 3 Experiments and Results 3.1 Feature Selection 4 Stacked Classifier 5 Conclusions 6 Aknowledgments Bibliographie Notes de bas de page Auteurs

    Texte intégral

    1 Introduction

    1Native Language Identification (NLI) is the research topic aimed at identifying the native language (L1) of a speaker or a writer based on his/her language production in a non-native language (L2). The leading assumption of NLI research is that speakers with the same L1 exhibit similar linguistic patterns in their L2 productions which can be viewed as traces of the L1 interference phenomena. Thanks to the availability of large-scale benchmark corpora, such as the TOEFL11 corpus (Blanchard et al., 2013), NLI has been recently gaining attention also in the NLP community where it is mainly addressed as a multi-class supervised classification task. This is the approach followed by the more recent systems taking part to the last editions of the NLI Shared Tasks held in 2013 (Tetreault et al., 2013) and 2017 (Malmasi et al., 2017). Typically, these systems exploit a variety of features encoding the linguistic structure of L2 text in terms of e.g. n-grams of characters, words, POS tags, syntactic constructions. Such features are used as input for machine learning algorithms, mostly based on traditional Support Vector Machine (SVM) models. In addition, rather than using the output of a single classifier, the most effective approach relies on ensemble methods based on multiple classifiers (Malmasi and Dras, 2017).

    2In this paper we want to further contribute to NLI research by focusing the attention on the role played by different types of linguistic features in predicting the native language of L2 writers. Starting from the approach devised by (Cimino and Dell'Orletta, 2017), which obtained the first position in the essay track of the 2017 NLI Shared Task, we carry out a systematic feature selection analysis to identify which features are more effective to capture traces of the native language in L2 writings at sentence and document level.

    3Our Contributions (i) We introduce for the first time a NLI sentence classification scenario, reporting the classification results; (ii) We study which features among a wide set of features contribute more to the sentence and to the document classification task; (iii) We investigate the contribution of features extracted from the sentence classifier in a stacked sentence-document system.

    2 The Classifier and Features

    4In this work, we built a classifier based on SVM using LIBLINEAR (Rong-En et al., 2008) as machine learning library. The set of documents described in Section 3 was automatically POS tagged by the part–of–speech tagger described in (Cimino and Dell'Orletta, 2016) and dependency–parsed by DeSR (Attardi et al., 2009). A wide set of features was considered in the classification of both sentences and documents. As shown in Table 1, they span across multiple levels of linguistic analysis. These features and the classifier were chosen since they were used by the 1st ranked classification system (Cimino and Dell'Orletta, 2017) in the 2017 NLI shared task.

    Table 1: Features used for document and sentence classification (* only for document)

    Raw text features

    TOEFL11 essay prompt*

    Text length (n. of tokens)

    Word length (avg. n. of characters)

    Average sentence length and standard deviation*

    Character n-grams (up to 8)

    Word n-grams (up to 4)

    Functional word n-grams (up to 3)

    Lemma n-grams (up to 4)

    Lexical features

    Type/token ratio of the first 100, 200, 300, 400 tokens*

    Etymological WordNet features [Melo2014]

    etymological n-grams (up-to 4)

    Morpho–syntactic features

    Coarse Part-Of-Speech n-grams (up to 4)

    Coarse Part-Of-Speech+Lemma of the following token n-grams (up to 4)

    Syntactic features

    Dependency type n-grams (sentence linear order) (up to 4)

    Dependency type n-grams (syntactic hierarchical order) (up to 4)

    Dependency subtrees (dependency of a word + the dependencies to its siblings in the sentence linear order)

    3 Experiments and Results

    5We carried out two experiments devoted to classify L2 documents and sentences. The training and development set distributed in the 2017 NLI shared task, i.e. the TOEFL11 corpus (Blanchard et al., 2013), was used as training data. It includes 12,100 documents, corresponding to a total of 198,334 sentences. The experiments were tested on the 2017 test set, including 1,100 documents (18,261 sentences).

    6The obtained macro average F1-scores were: 0.8747 in the document classification task and 0.4035 in the sentence one. As it was expected, the identification of the L1 of the sentences turned out as a more complex task than L1 document classification. Both document and sentence classification are influenced by the number of words but with a different impact. Figure 1 shows that the average performance on sentences is reached for sentences ~21–token long, which corresponds to the average sentence length for this dataset. As the sentence length increases, the accuracy increases as well. Due to the smaller amount of linguistic evidence, the classification of short sentences is a more complex task. The performance of document classification is more stable: the best f–score is already reached for documents of ~200–tokens, which corresponds to a very short document compared to the average size of TOEFL11 documents (330 tokens).

    Figure 1: Sentence classification performance across bins of sentences of the same lengths

    Image 100000000000038600000222F915E127607CBCC7.jpg

    7Figures 2(a) and 2(b) report the confusion matrices of the two experiments1. As it can be seen, both for sentences and documents the best classification performance is obtained for German, Japanese and Chinese, even though with some differences in the relative ranking positions, e.g. German is the top ranked one in the sentence classification scenario and the 2nd ranked one in the document classification one, while Japanese is the best classified L1 in the document experiment and the 4th ranked one in the sentence classification scenario. Conversely, we observe differences with respect to the worst recognized L1s, which are Turkish, Hindi and Korean in the document classification task and Arabic, Spanish and Turkish in the sentence classification one. The two confusion matrices also reveal a peculiar error distribution trend: the confusion matrix of the sentence classification model is much more sparse than the document classification one. This means that for each considered L1, the errors made by the sentence classifier are quite similarly distributed over all possible L1s; instead, errors in the document classification scenario are much more prototypical, i.e. the wrong predicted label is assigned to only one or two L1 candidates, which change according to the specific L1. This is shown e.g. by languages belonging to same language family such as Japanese and Korean which belong to the same Altaic family. Specifically, in the document classification scenario Korean is mainly confused with Japanese (10% of errors). This trend holds also in the sentence classification experiment where 17.8% of errors were due to the confusion of Korean with Japanese and vice versa (18.2% of errors). Interestingly enough, the most prototypical errors were also made when contact languages were concerned. This is for example the case of Hindi and Telugu: Hindi documents were mainly confused with Telugu ones (16% of errors) and Telugu documents with Hindi ones (13% of errors). Similarly, in the sentence classification scenario, Hindi sentences were wrongly classified as Telugu sentences in about 20% of cases and vice versa. As previously shown by Cimino et al. (2013), even if these two languages do not belong to the same family, such classification errors might originate from a similar linguistic profile due to language contact phenomena: for instance, both Hindi and Telugu L1 essays are characterized by sentences and words of similar length, or they share similar syntactic structures such e.g. parse trees of similar depth and embedded complement chains governed by a nominal head of similar length.

    Figure 2: Confusion matrix of document (a) and sentence classification (b)

    Image 1000000000000638000002712122CD95CCD0CF13.jpg

    8The behavior of the two classifiers may suggest that i) some features could play a different role in the classification of sentences with respect to documents and ii) the document classifier can be improved using features extracted from the output of a sentence classifier in a stacked configuration. To investigate these hypotheses, we carried out an extensive feature selection analysis to study the role of the features in the two classification scenarios.

    3.1 Feature Selection

    9In the first step of the feature selection process, we extracted all the features from the training set and pruned those occurring less than 4 times, obtaining ~ 4,000,000 distinct features both for document and sentence classification. In the second step, we ranked the extracted features through the Recursive Feature Elimination (RFE) algorithm implemented in the Scikit-learn library al.[et using Linear SVM as estimator algorithm. We dropped 1% of features in each iteration. At the end of this step we selected the top ranked features corresponding to ~ 40,000 features both for the sentence and document tasks. These features were further re-ranked using the RFE algorithm (dropping 100 features at each iteration) to allow a more fine grained analysis.

    10Figure 3(a) compares the percentage of different types of features used in the classification of documents and sentences. As it can be noted, the document classifier uses more words n-grams, especially n-grams characters. Instead, morpho–syntactic and syntactic features are more effective for sentence classification, and the n-grams of lemmas even more than 4 times. Figures 3(b), 3(c) and 3(d) show the variation of relevance of the 40k raw text, morpho-syntactic and syntactic features grouped in bins of 100 features. The lines in the charts correspond to the differences between document and sentence in terms of percentage of a single type of feature in the bin with respect to its total distribution in the whole 40k selected features2. Negative values mean that this distribution in the bin is higher for sentence classification.

    Figure 3: Distribution of the first top ranked 40k features in the document and sentence classification

    Image 100000000000070800000516A06ED0A884298FA3.jpg

    11Among the raw text features (Figure 3(b)), n-grams of words occur more in the 1st bins of document classification, while n-grams of characters and lemma are more relevant in the 1st bins of sentence classification. The n-grams of coarse parts–of-speech are equally distributed in the two rankings, instead both the n-grams of coarse parts–of-speech followed by a lemma and the n-grams of functional words occur more in the 1st bins of sentence classification (Figure 3(c)). This confirms the key role played by lemma in sentence classification.

    12For what concerns syntactic information (Figure 3(d)), the features that properly capture sentence structure (dependency subtree and the hierarchical syntactic dependencies) are all contained in the first bins of document classification even if their total distribution is lower than in the sentence. This shows that syntactic information is very relevant also when longer texts are classified and that this kind of information is not captured by n-grams of words. Feature types with low number of instances are not reported in these charts. Among these, etymological n-grams appears in the first bins both for sentence and document, confirming the relevance of the etymological information already proven for NLI document classification (Nastase and Strapparava, 2017). For sentence classification, it is also relevant sentence length and word length. Instead, for document, type/token ratio plays a very important role. Interestingly, the average sentence length does not appear in the 40k features; we found instead the sentence length standard deviation, showing that what counts more is the variation in length rather than the average value. Even though not contained in the first bins, also word and document lengths and the TOEFL11 essay prompt are in the top 40k features.

    4 Stacked Classifier

    13The different role of the features in the two L1 classification tasks suggests that we may improve the traditional NLI document classification by combining sentence and document classifiers. We thus evaluated and extended the stacked sentence-document architecture proposed by (Cimino and Dell'Orletta, 2017). In addition to the linguistic features, they proposed a stacked system using the L1 predictions of a pre–trained sentence classifier to train a document classifier. Thus we run several experiments on the NLI Shared Task 2017 test set to assess i) the importance of the sentence classifier in a stacked sentence-document architecture and ii) which features extracted from the predictions of the L1 sentence classifier maximize the accuracy of the stacked system. The sentence classifier assigned a confidence score for each L1 to each sentence of the documents. Based on the confidence score, we defined the following features: for each L1 i) the mean sentence confidence (avg), ii) the standard deviation of confidences (stddev), iii) the product of the confidences (prod), iv) the top–3 highest and lowest confidence values (top–3 max-min). The last two features were introduced to mitigate the effect of spike values that may be introduced by considering the max-min L1 confidences used in (Cimino and Dell'Orletta, 2017). The first row of Table 2 reports the result obtained by (Cimino and Dell'Orletta, 2017) by the stacked classifier on the same test set. The second row reports the results of our document system which does not use features extracted from the sentence classifier. The third row reports the result of a classifier that uses only the features extracted from the predictions of the L1 sentence classifier. The following rows report the contribution of each sentence classifier feature in the stacked architecture showing an improvement (with the exception of the product) with respect to the base classifier. The top–3 highest and lowest confidence values are the most helpful features in a stacked architecture. The best result is obtained when using all the sentence classifier features in the base classifier, which is the state-of-the-art on the 2017 NLI test set.

    Table 2: Results of the stacked system

    Model

    F1-Score

    Cimino and Dell’Orletta (2017)

    0.8818

    Base classifier

    0.8747

    Sentence features

    0.8363

    Base class. + avg

    0.8773

    Base class. + stddev

    0.8773

    Base class. + prod

    0.8747

    Base class. + top–3 max-min

    0.8800

    Base class. + all sentence feat.

    0.8828

    5 Conclusions

    14We introduced a new NLI scenario focused on sentence classification. Compared to document classification we obtained different results in terms of accuracy and distribution of errors across the L1s. We showed the different role played by a wide set of linguistic features in the two NLI scenarios. These differences may justify the performance boost we achieved with a stacked sentence-document system. We also assessed which features extracted from the sentence classifier maximizes NLI document classification.

    6 Aknowledgments

    15The work presented in this paper was partially supported by the 2–year project (2018-2020) SchoolChain – Soluzioni innovative per la creazione, la certificazione, il riuso e la condivisione di unità didattiche digitali all’interno del sistema Scuola, funded by Regione Toscana (BANDO POR FESR 2014-2020).

    Bibliographie

    Giuseppe Attardi, Felice Dell’Orletta, Maria Simi, and Joseph Turian. 2009. Accurate dependency parsing with a stacked multilayer perceptron. In Proceedings of the 2nd Workshop of Evalita 2009. December, Reggio Emilia, Italy.

    Daniel Blanchard, Joel Tetreault, Derrick Higgins, Aoife Cahill, and Martin Chodorow. 2013. TOEFL11: A Corpus of Non-Native English. Technical report, Educational Testing Service.

    Andrea Cimino and Felice Dell’Orletta. 2016. Building the state-of-the-art in POS tagging of italian tweets. In Proceedings of Third Italian Conference on Computational Linguistics (CLiC-it 2016) & Fifth Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop (EVALITA), December 5-7.

    Andrea Cimino and Felice Dell’Orletta. 2017. Stacked sentence–document classifier approach for improving native language identification. In Proceedings of the 12th Workshop on Innovative Use of NLP for Building Educational Applications, BEA@EMNLP 2017, September 8, pages 430-437.

    Andrea Cimino, Felice Dell’Orletta, Giulia Venturi, and Simonetta Montemagni. 2013. Linguistic Profiling based on General-purpose Features and Native Language Identification. In Proceedings of the Eighth Workshop on Innovative Use of NLP for Building Educational Applications, BEA@NAACL-HLT 2013, June 13, 2013, Atlanta, Georgia, USA, pages 207–215.

    Shervin Malmasi and Mark Dras. 2017. Native Language Identification using Stacked Generalization. arXiv preprint arXiv:1703.06541.

    Shervin Malmasi, Keelan Evanini, Aoife Cahill, Joel Tetreault, Robert Pugh, Christopher Hamill, Diane Napolitano and Yao Qian 2017. A Report on the 2017 Native Language Identification Shared Task. In Proceedings of the 12th Workshop on Building Educational Applications Using NLP. BEA@EMNLP 2017, September 8.

    Gerard de Melo. 2014. Etymological Wordnet: Tracing the history of words. In Proceedings of the 9th Language Resources and Evaluation Conference (LREC 2014), Paris, France. ELRA.

    Vivi Nastase and Carlo Strapparava. 2017. Word etymology as native language interference. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 2702–2707.

    F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot, and E. Duchesnay. 2011. Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12:2825–2830.

    Fan Rong-En, Chang Kai-Wei, Hsieh Cho-Jui, Wang Xiang-Rui, and Lin Chih-Jen. 2008. LIBLINEAR: A library for large linear classification. 2008. LIBLINEAR: A library for large linear classification. Journal of Machine Learning Research, 9:1871–1874.

    Joel Tetreault, Daniel Blanchard, and Aoife Cahill. 2013. A Report on the First Native Language Identification Shared Task. In Proceedings of the Eighth Workshop on Innovative Use of NLP for Building Educational Applications, Atlanta, GA, USA. Association for Computational Linguistics.

    Notes de bas de page

    1 Since the number of documents and sentences in the two experiments is different, in order to make comparable the values of the two confusion matrices, the sentence classification values were normalized to 100.

    2 Spline interpolation applied for readability purpose.

    Auteurs

    • Andrea Cimino

      Istituto di Linguistica Computazionale “Antonio Zampolli” (ILC–CNR), ItaliaNLP Lab - www.italianlp.it – andrea.cimino[at]ilc.cnr.it

    • Felice Dell’Orletta

      Istituto di Linguistica Computazionale “Antonio Zampolli” (ILC–CNR), ItaliaNLP Lab - www.italianlp.it – felice.dellorletta[at]ilc.cnr.it

    • Dominique Brunato

      Istituto di Linguistica Computazionale “Antonio Zampolli” (ILC–CNR), ItaliaNLP Lab - www.italianlp.it – dominique.brunato[at]ilc.cnr.it

    • Giulia Venturi

      Istituto di Linguistica Computazionale “Antonio Zampolli” (ILC–CNR), ItaliaNLP Lab - www.italianlp.it – giulia.venturi[at]ilc.cnr.it

    Précédent Suivant
    Table des matières

    Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

    Voir plus de livres
    Proceedings of the Second Italian Conference on Computational Linguistics CLiC-it 2015

    Proceedings of the Second Italian Conference on Computational Linguistics CLiC-it 2015

    3-4 December 2015, Trento

    Cristina Bosco, Sara Tonelli et Fabio Massimo Zanzotto (dir.)

    2015

    Proceedings of the Third Italian Conference on Computational Linguistics CLiC-it 2016

    Proceedings of the Third Italian Conference on Computational Linguistics CLiC-it 2016

    5-6 December 2016, Napoli

    Anna Corazza, Simonetta Montemagni et Giovanni Semeraro (dir.)

    2016

    EVALITA. Evaluation of NLP and Speech Tools for Italian

    EVALITA. Evaluation of NLP and Speech Tools for Italian

    Proceedings of the Final Workshop 7 December 2016, Naples

    Pierpaolo Basile, Franco Cutugno, Malvina Nissim et al. (dir.)

    2016

    Proceedings of the Fourth Italian Conference on Computational Linguistics CLiC-it 2017

    Proceedings of the Fourth Italian Conference on Computational Linguistics CLiC-it 2017

    11-12 December 2017, Rome

    Roberto Basili, Malvina Nissim et Giorgio Satta (dir.)

    2017

    Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-it 2018

    Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-it 2018

    10-12 December 2018, Torino

    Elena Cabrio, Alessandro Mazzei et Fabio Tamburini (dir.)

    2018

    EVALITA Evaluation of NLP and Speech Tools for Italian

    EVALITA Evaluation of NLP and Speech Tools for Italian

    Proceedings of the Final Workshop 12-13 December 2018, Naples

    Tommaso Caselli, Nicole Novielli, Viviana Patti et al. (dir.)

    2018

    EVALITA Evaluation of NLP and Speech Tools for Italian - December 17th, 2020

    EVALITA Evaluation of NLP and Speech Tools for Italian - December 17th, 2020

    Proceedings of the Seventh Evaluation Campaign of Natural Language Processing and Speech Tools for Italian Final Workshop

    Valerio Basile, Danilo Croce, Maria Maro et al. (dir.)

    2020

    Proceedings of the Seventh Italian Conference on Computational Linguistics CLiC-it 2020

    Proceedings of the Seventh Italian Conference on Computational Linguistics CLiC-it 2020

    Bologna, Italy, March 1-3, 2021

    Felice Dell'Orletta, Johanna Monti et Fabio Tamburini (dir.)

    2020

    Proceedings of the Eighth Italian Conference on Computational Linguistics CliC-it 2021

    Proceedings of the Eighth Italian Conference on Computational Linguistics CliC-it 2021

    Milan, Italy, 26-28 January, 2022

    Elisabetta Fersini, Marco Passarotti et Viviana Patti (dir.)

    2022

    Voir plus de livres
    1 / 9
    Proceedings of the Second Italian Conference on Computational Linguistics CLiC-it 2015

    Proceedings of the Second Italian Conference on Computational Linguistics CLiC-it 2015

    3-4 December 2015, Trento

    Cristina Bosco, Sara Tonelli et Fabio Massimo Zanzotto (dir.)

    2015

    Proceedings of the Third Italian Conference on Computational Linguistics CLiC-it 2016

    Proceedings of the Third Italian Conference on Computational Linguistics CLiC-it 2016

    5-6 December 2016, Napoli

    Anna Corazza, Simonetta Montemagni et Giovanni Semeraro (dir.)

    2016

    EVALITA. Evaluation of NLP and Speech Tools for Italian

    EVALITA. Evaluation of NLP and Speech Tools for Italian

    Proceedings of the Final Workshop 7 December 2016, Naples

    Pierpaolo Basile, Franco Cutugno, Malvina Nissim et al. (dir.)

    2016

    Proceedings of the Fourth Italian Conference on Computational Linguistics CLiC-it 2017

    Proceedings of the Fourth Italian Conference on Computational Linguistics CLiC-it 2017

    11-12 December 2017, Rome

    Roberto Basili, Malvina Nissim et Giorgio Satta (dir.)

    2017

    Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-it 2018

    Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-it 2018

    10-12 December 2018, Torino

    Elena Cabrio, Alessandro Mazzei et Fabio Tamburini (dir.)

    2018

    EVALITA Evaluation of NLP and Speech Tools for Italian

    EVALITA Evaluation of NLP and Speech Tools for Italian

    Proceedings of the Final Workshop 12-13 December 2018, Naples

    Tommaso Caselli, Nicole Novielli, Viviana Patti et al. (dir.)

    2018

    EVALITA Evaluation of NLP and Speech Tools for Italian - December 17th, 2020

    EVALITA Evaluation of NLP and Speech Tools for Italian - December 17th, 2020

    Proceedings of the Seventh Evaluation Campaign of Natural Language Processing and Speech Tools for Italian Final Workshop

    Valerio Basile, Danilo Croce, Maria Maro et al. (dir.)

    2020

    Proceedings of the Seventh Italian Conference on Computational Linguistics CLiC-it 2020

    Proceedings of the Seventh Italian Conference on Computational Linguistics CLiC-it 2020

    Bologna, Italy, March 1-3, 2021

    Felice Dell'Orletta, Johanna Monti et Fabio Tamburini (dir.)

    2020

    Proceedings of the Eighth Italian Conference on Computational Linguistics CliC-it 2021

    Proceedings of the Eighth Italian Conference on Computational Linguistics CliC-it 2021

    Milan, Italy, 26-28 January, 2022

    Elisabetta Fersini, Marco Passarotti et Viviana Patti (dir.)

    2022

    Voir plus de chapitres

    ISACCO: a corpus for investigating spoken and written language evelopment in Italian school–age children

    Dominique Brunato et Felice Dell’Orletta

    Identifying Predictive Features for Textual Genre Classification: the Key Role of Syntax

    Andrea Cimino, Martijn Wieling, Felice Dell’Orletta et al.

    Multi-task Learning in Deep Neural Networks at EVALITA 2018

    Andrea Cimino, Lorenzo De Mattei et Felice Dell’Orletta

    Building the state-of-the-art in POS tagging of Italian Tweets

    Andrea Cimino et Felice Dell’Orletta

    TAG-it @ EVALITA2020: Overview of the Topic, Age, and Gender Prediction Task for Italian

    Andrea Cimino, Felice Dell’Orletta et Malvina Nissim

    Tandem LSTM-SVM Approach for Sentiment Analysis

    Andrea Cimino et Felice Dell’Orletta

    Overview of the EVALITA 2018 Cross-Genre Gender Prediction (GxG) Task

    Felice Dell’Orletta et Malvina Nissim

    AcCompl-it @ EVALITA2020: Overview of the Acceptability & Complexity Evaluation Task for Italian

    Dominique Brunato, Cristiano Chesi, Felice Dell’Orletta et al.

    DARC-IT: a DAtaset for Reading Comprehension in Italian

    Dominique Brunato, Martina Valeriani et Felice Dell’Orletta

    Quantitative Linguistic Investigations across Universal Dependencies Treebanks

    Chiara Alzetta, Felice Dell’Orletta, Simonetta Montemagni et al.

    Tracking the Evolution of Written Language Competence: an NLP–based Approach

    Stefan Richter, Andrea Cimino, Felice Dell’Orletta et al.

    Voir plus de chapitres
    1 / 11

    ISACCO: a corpus for investigating spoken and written language evelopment in Italian school–age children

    Dominique Brunato et Felice Dell’Orletta

    Identifying Predictive Features for Textual Genre Classification: the Key Role of Syntax

    Andrea Cimino, Martijn Wieling, Felice Dell’Orletta et al.

    Multi-task Learning in Deep Neural Networks at EVALITA 2018

    Andrea Cimino, Lorenzo De Mattei et Felice Dell’Orletta

    Building the state-of-the-art in POS tagging of Italian Tweets

    Andrea Cimino et Felice Dell’Orletta

    TAG-it @ EVALITA2020: Overview of the Topic, Age, and Gender Prediction Task for Italian

    Andrea Cimino, Felice Dell’Orletta et Malvina Nissim

    Tandem LSTM-SVM Approach for Sentiment Analysis

    Andrea Cimino et Felice Dell’Orletta

    Overview of the EVALITA 2018 Cross-Genre Gender Prediction (GxG) Task

    Felice Dell’Orletta et Malvina Nissim

    AcCompl-it @ EVALITA2020: Overview of the Acceptability & Complexity Evaluation Task for Italian

    Dominique Brunato, Cristiano Chesi, Felice Dell’Orletta et al.

    DARC-IT: a DAtaset for Reading Comprehension in Italian

    Dominique Brunato, Martina Valeriani et Felice Dell’Orletta

    Quantitative Linguistic Investigations across Universal Dependencies Treebanks

    Chiara Alzetta, Felice Dell’Orletta, Simonetta Montemagni et al.

    Tracking the Evolution of Written Language Competence: an NLP–based Approach

    Stefan Richter, Andrea Cimino, Felice Dell’Orletta et al.

    Accès ouvert

    Accès ouvert

    ePub

    PDF

    PDF du chapitre

    1 Since the number of documents and sentences in the two experiments is different, in order to make comparable the values of the two confusion matrices, the sentence classification values were normalized to 100.

    2 Spline interpolation applied for readability purpose.

    Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-it 2018

    X Facebook Email

    Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-it 2018

    Ce livre est cité par

    • Leoni, Chiara. Coccoli, Mauro. Torre, Ilaria. Vercelli, Gianni. (2020) Lecture Notes in Computer Science Intelligent Data Engineering and Automated Learning – IDEAL 2020. DOI: 10.1007/978-3-030-62365-4_34
    • Artese, Maria Teresa. Ciocca, Gianluigi. Gagliardi, Isabella. (2021) Lecture Notes in Computer Science Pattern Recognition. ICPR International Workshops and Challenges. DOI: 10.1007/978-3-030-68821-9_53
    • Konat, Barbara. Gajewska, Ewelina. Rossa, Wiktoria. (2024) Pathos in Natural Language Argumentation: Emotional Appeals and Reactions. Argumentation, 38. DOI: 10.1007/s10503-024-09631-2
    • Konat, Barbara. (2025) Argumentation Library Emotional Appeals in Argumentation. DOI: 10.1007/978-3-032-06332-8_8

    Ce chapitre est cité par

    • del Río, Iria. (2020) Lecture Notes in Computer Science Computational Processing of the Portuguese Language. DOI: 10.1007/978-3-030-41505-1_9
    • Mattei, Andrea. Brunato, Dominique. Dell’Orletta, Felice. (2020) Proceedings of the Seventh Italian Conference on Computational Linguistics CLiC-it 2020. DOI: 10.4000/books.aaccademia.8718
    • Cerulli, Aldo. Brunato, Dominique. Dell’Orletta, Felice. (2023) Linguistic Profile of a Text and Human Ratings of Writing Quality: a Case Study on Italian L1 Learner Essays. Italian Journal of Computational Linguistics, 9. DOI: 10.4000/ijcol.1104

    Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-it 2018

    Vous pouvez vous connecter à votre bibliothèque à l’adresse suivante : https://freemium.openedition.org/oebooks

    Suggérer l’acquisition à votre bibliothèque

    Si vous avez des questions, vous pouvez nous écrire à access[at]openedition.org

    Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-it 2018

    Vérifiez si votre bibliothèque a déjà acquis ce livre : authentifiez-vous à OpenEdition Freemium for Books.

    Vous pouvez suggérer à votre bibliothèque d’acquérir un ou plusieurs livres publiés sur OpenEdition Books. N’hésitez pas à lui indiquer nos coordonnées : access[at]openedition.org

    Vous pouvez également nous indiquer, à l’aide du formulaire suivant, les coordonnées de votre bibliothèque afin que nous la contactions pour lui suggérer l’achat de ce livre. Les champs suivis de (*) sont obligatoires.

    Veuillez, s’il vous plaît, remplir tous les champs.

    La syntaxe de l’email est incorrecte.

    Référence numérique du chapitre

    Format

    Cimino, A., Dell’Orletta, F., Brunato, D., & Venturi, G. (2018). Sentences and Documents in Native Language Identification. In E. Cabrio, A. Mazzei, & F. Tamburini (éds.), Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-it 2018. Torino: Accademia University Press. https://doi.org/10.4000/books.aaccademia.3204
    Cimino, Andrea, Felice Dell’Orletta, Dominique Brunato, et Giulia Venturi. « Sentences and Documents in Native Language Identification ». In Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-It 2018, édité par Elena Cabrio, Alessandro Mazzei, et Fabio Tamburini. Torino: Accademia University Press, 2018. doi:10.4000/books.aaccademia.3204.
    Cimino, Andrea, et al. « Sentences and Documents in Native Language Identification ». Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-It 2018, édité par Elena Cabrio et al., Accademia University Press, 2018, https://doi.org/10.4000/books.aaccademia.3204.

    Référence numérique du livre

    Format

    Cabrio, E., Mazzei, A., & Tamburini, F. (éds.). (2018). Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-it 2018. Torino: Accademia University Press. https://doi.org/10.4000/books.aaccademia.2802
    Cabrio, Elena, Alessandro Mazzei, et Fabio Tamburini, éd. Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-It 2018. Torino: Accademia University Press, 2018. doi:10.4000/books.aaccademia.2802.
    Cabrio, Elena, et al., éditeurs. Proceedings of the Fifth Italian Conference on Computational Linguistics CLiC-It 2018. Accademia University Press, 2018, https://doi.org/10.4000/books.aaccademia.2802.
    Compatible avec Zotero Zotero

    1 / 3

    Accademia University Press

    Accademia University Press

    • Plan du site
    • Se connecter

    Suivez-nous

    • Facebook
    • Flux RSS

    URL : http://www.aaccademia.it/

    Email : info@aaccademia.it

    Adresse :

    Accademia University Press

    Via Carlo Alberto 55

    I‐10123

    Torino

    Italia

    OpenEdition
    • Candidater à OpenEdition Books
    • Connaître le programme OpenEdition Freemium
    • Commander des livres
    • S’abonner à la lettre d’OpenEdition
    • CGU d’OpenEdition Books
    • Accessibilité : partiellement conforme
    • Données personnelles
    • Gestion des cookies
    • Système de signalement