Papers reviewed and determined not to be word norm studies. Use the flag icon to report errors or suggest re-inclusion.
18625 papers
Processing unpleasant affective cues induces elevated momentary symptom reports, especially in persons with high levels of symptom reporting in daily life. The present study aimed to examine whether applying an emotion regulation strategy, i.e. affect labeling, can inhibit these emotion influences on symptom reporting. Student participants (N = 61) with varying levels of habitual symptom reporting completed six picture viewing trials of homogeneous valence (three pleasant, three unpleasant) under three conditions: merely viewing, emotional labeling, or content (non-emotional) labeling. Affect ratings and symptom reports were collected after each trial. Participants completed a motor inhibition task and self-control questionnaires as indices of their inhibitory capacities. Heart rate variability was also measured. Labeling, either emotional or non-emotional, significantly reduced experienced affect, as well as the elevated symptoms reports observed after unpleasant picture viewing. These labeling effects became more pronounced with increasing levels of habitual symptom reporting, suggesting a moderating role of the latter variable, but did not correlate with any index of general inhibitory capacity. Our findings suggest that using an emotion regulation strategy, such as labeling emotional stimuli, can reverse the effects of unpleasant stimuli on symptom reporting and that such strategies can be especially beneficial for individuals suffering from medically unexplained physical symptoms.
The increasing diversity of languages used on the web introduces a new level of complexity to Information Retrieval (IR) systems. We can no longer assume that textual content is written in one language or even the same language family. In this paper, we demonstrate how to build massive multilingual annotators with minimal human expertise and intervention. We describe a system that builds Named Entity Recognition (NER) annotators for 40 major languages using Wikipedia and Freebase. Our approach does not require NER human annotated datasets or language specific resources like treebanks, parallel corpora, and orthographic rules. The novelty of approach lies therein - using only language agnostic techniques, while achieving competitive performance. Our method learns distributed word representations (word embeddings) which encode semantic and syntactic features of words in each language. Then, we automatically generate datasets from Wikipedia link structure and Freebase attributes. Finally, we apply two preprocessing stages (oversampling and exact surface form matching) which do not require any linguistic expertise. Our evaluation is two fold: First, we demonstrate the system performance on human annotated datasets. Second, for languages where no gold-standard benchmarks are available, we propose a new method, distant evaluation, based on statistical machine translation.
We develop an instance (token) based extension of the state of the art word (type) based part-ofspeech induction system introduced in (Yatbaz et al., 2012). Each word instance is represented by a feature vector that combines information from the target word and probable substitutes sampled from an n-gram model representing its context. Modeling ambiguity using an instance based model does not lead to significant gains in overall accuracy in part-of-speech tagging because most words in running text are used in their most frequent class (e.g. 93.69% in the Penn Treebank). However it is important to model ambiguity because most frequent words are ambiguous and not modeling them correctly may negatively affect upstream tasks. Our main contribution is to show that an instance based model can achieve significantly higher accuracy on ambiguous words at the cost of a slight degradation on unambiguous ones, maintaining a comparable overall accuracy. On the Penn Treebank, the overall many-to-one accuracy of the system is within 1% of the state-of-the-art (80%), while on highly ambiguous words it is up to 70% better. On multilingual experiments our results are significantly better than or comparable to the best published word or instance based systems on 15 out of 19 corpora in 15 languages. The vector representations for words used in our system are available for download for further experiments.
Recent years have seen an increased interest in and availability of parallel corpora. Large corpora from international organizations (e.g. European Union, United Nations, European Patent Office), or from multilingual Internet sites (e.g. OpenSubtitles) are now easily available and are used for statistical machine translation but also for online search by different user groups. This paper gives an overview of different usages and different types of search systems. In the past, parallel corpus search systems were based on sentence-aligned corpora. We argue that automatic word alignment allows for major innovations in searching parallel corpora. Some online query systems already employ word alignment for sorting translation variants, but none supports the full query functionality that has been developed for parallel treebanks. We propose to develop such a system for efficiently searching large parallel corpora with a powerful query language.
The feminist movement purports to improve conditions for women, and yet only a minority of women in modern societies self-identify as feminists. This is known as the feminist paradox. It has been suggested that feminists exhibit both physiological and psychological characteristics associated with heightened masculinization, which may predispose women for heightened competitiveness, sex-atypical behaviors, and belief in the interchangeability of sex roles. If feminist activists, i.e., those that manufacture the public image of feminism, are indeed masculinized relative to women in general, this might explain why the views and preferences of these two groups are at variance with each other. We measured the 2D:4D digit ratios (collected from both hands) and a personality trait known as dominance (measured with the Directiveness scale) in a sample of women attending a feminist conference. The sample exhibited significantly more masculine 2D:4D and higher dominance ratings than comparison samples representative of women in general, and these variables were furthermore positively correlated for both hands. The feminist paradox might thus to some extent be explained by biological differences between women in general and the activist women who formulate the feminist agenda.
We present an algorithm and implementation for extracting recurring fragments from treebanks. Using a tree-kernel method the largest common fragments are extracted from each pair of trees. The algorithm presented achieves a thirty-fold speedup over the previously available method on the Wall Street Journal dataset. It is also more general, in that it supports trees with discontinuous constituents. The resulting fragments can be used as a tree-substitution grammar or in classification problems such as authorship attribution and other stylometry tasks.
En este artículo se describe el proceso que se siguió para crear el componente anotado con información lingüística (treebank) del Corpus de Mensajes Presidenciales Costarricenses (CODIMEP-CR), en el marco del proyecto No. 745-B1-244 Interfaz para el procesamiento de corpus lingüísticos digitales-IPROCOLDI. Ambos corpus se albergan en la interfaz IPROCOLDI (http://163.178.116.145/iprocoldi/).
One way of teaching grammar, namely morphology and syntax, is to visualize sentences as diagrams capturing relationships between words. Similarly, such relationships are captured in a more complex way in treebanks serving as key building stones in modern natural language processing. However, building them is very time consuming, thus we have been seeking for an alternative cheaper and faster way, like crowdsourcing. The purpose of our work is to explore possibility to get sentence diagrams produced by students and teachers. In our pilot study, the object language is Czech, where sentence diagrams are part of elementary school curriculum.
Automatic text categorisation systems is a type of software that every day it is receiving more interest, due not only to its use in documentaries environments but also to its possible application to tag properly documents on the Web. Many options have been proposed to face this subject using statistical approaches, natural language processing tools, ontologies and lexical databases. Nevertheless, there have been no too many empirical evaluations comparing the influence of the different tools used to solve these problems, particularly in a multilingual environment. In this paper we propose a multi-language rule-based pipeline system for automatic document categorisation and we compare empirically the results of applying techniques that rely on statistics and supervised learning with the results of applying the same techniques but with the support of smarter tools based on language semantics and ontologies, using for this purpose several corpora of documents. GENIE is being applied to real environments, which shows the potential of the proposal.
We investigate whether parsers can be used for self-monitoring in surface realization in order to avoid egregious errors involving "vicious" ambiguities, namely those where the intended interpretation fails to be considerably more likely than alternative ones. Using parse accuracy in a simple reranking strategy for selfmonitoring, we find that with a stateof-the-art averaged perceptron realization ranking model, BLEU scores cannot be improved with any of the well-known Treebank parsers we tested, since these parsers too often make errors that human readers would be unlikely to make. However, by using an SVM ranker to combine the realizer's model score together with features from multiple parsers, including ones designed to make the ranker more robust to parsing mistakes, we show that significant increases in BLEU scores can be achieved. Moreover, via a targeted manual analysis, we demonstrate that the SVM reranker frequently manages to avoid vicious ambiguities, while its ranking errors tend to affect fluency much more often than adequacy.
The article is devoted to the problem of translation from French into Polish from the perspective of the object oriented approach proposed by Wiesław Banyś. The author takes into consideration some of the problems, both of theoretical and practical nature, which appear while working on the formation of contrastive lexical database for automatic translation of texts. While analyzing specific chosen examples which may cause different kinds of problems in the description, the author offers their interpretations in the target language in accordance with the adopted approach.
The Stuttgart-Tbingen Tagset (STTS) is a widely used POS annotation scheme for German which provides 54 different tags for the analysis on the part of speech level.The tagset, however, does not distinguish between adverbs and different types of particles used for expressing modality, intensity, graduation, or to mark the focus of the sentence.In the paper, we present an extension to the STTS which provides tags for a more fine-grained analysis of modification, based on a syntactic perspective on parts of speech.We argue that the new classification not only enables us to do corpus-based linguistic studies on modification, but also improves statistical parsing.We give proof of concept by training a data-driven dependency parser on data from the TiGer treebank, providing the parser a) with the original STTS tags and b) with the new tags.Results show an improved labelled accuracy for the new, syntactically motivated classification.
Constituent Context Model (CCM) is an effective generative model for grammar induction, the aim of which is to induce hierarchical syntactic structure from natural text. The CCM simply defines the Multinomial distribution over constituents, which leads to a severe data sparse problem because long constituents are unlikely to appear in unseen data sets. This paper proposes a Bayesian method for constituent smoothing by defining two kinds of prior distributions over constituents: the Dirichlet prior and the Pitman-Yor Process prior. The Dirichlet prior functions as an additive smoothing method, and the PYP prior functions as a back-off smoothing method. Furthermore, a modified CCM is proposed to differentiate left constituents and right constituents in binary branching trees. Experiments show that both the proposed Bayesian smoothing method and the modified CCM are effective, and combining them attains or significantly improves the state-of-the-art performance of grammar induction evaluated on standard treebanks of various languages.
Comunicació presentada al 9th International Conference on Language Resources and Evaluation (LREC'14), celebrat del 26 al 31 de maig de 2014 a Reykjavík, Islàndia.
English. Network theory provides a suitable framework to model the structure of language as a complex system. Based on a network built from a Latin dependency treebank, this paper applies methods for network analysis to show the key role of the verb sum (to be) in the overall structure of the network. Italiano. La teoria dei grafi fornisce un valido supporto alla modellizzazione strutturale del sistema linguistico. Basandosi su un network costruito a partire da una treebank a dipendenze del latino, l’articolo applica diversi metodi di analisi dei grafi, mostrando l’importanza del ruolo rivestito dal verbo sum (essere) nella struttura complessiva del network.
This paper mainly introduced the research on constructing Mongolian Treebank based on phrase structure grammar. Having Considered related Mongolian Treebank work and Mongolian words characteristics, we developed a Mongolian syntactic tagset. The tagset includes two kinds of tags. One is syntactic constituent tag and the other is grammatical relation tag. On the basis of the tagset, we developed the Mongolian Treebank auxiliary processing system. Finally, we built a Treebank that contains 3645 sentences and did an experiment on this Treebank.
In this paper, we analyze the impact of various dependency representations for various constructions on the general parsing accuracy and on the parsing accuracy of these constructions. We focus on the analysis of coordination constructions, complex predicates, and punctuation mark attachment. We use Latvian Treebank as a dataset, thus, providing insight for an inflective language with a rather free word order. Experiments with MaltParser, a transition-based parser, show clear difference in learnability of various representations for the considered constructions. Future work would include carrying out comparable experiments with a graph-based dependency parser like MSTParser.
According to Tsinghua Chinese Treebank annotation methods, the authors extracted relation words and marked their categories. Then syntax, lexical and position features of automatic syntax tree with and without functional marker were extracted to recognize and classify relation words. Experiment results show that relative recognition accuracy is 95.7%, and relation words classification F1 is 77.2%.
We describe a new dependency parser for English tweets, TWEEBOPARSER. The parser builds on several contributions: new syntactic annotations for a corpus of tweets (TWEEBANK), with conventions informed by the domain; adaptations to a statistical parsing algorithm; and a new approach to exploiting out-of-domain Penn Treebank data. Our experiments show that the parser achieves over 80% unlabeled attachment accuracy on our new, high-quality test set and measure the benefit of our contributions. Our dataset and parser can be found at http://www.ark.cs.cmu.edu/TweetNLP.
Recent work on Chinese analysis has led to large-scale annotations of the internal structures of words, enabling characterlevel analysis of Chinese syntactic structures. In this paper, we investigate the problem of character-level Chinese dependency parsing, building dependency trees over characters. Character-level information can benefit downstream applications by offering flexible granularities for word segmentation while improving wordlevel dependency parsing accuracies. We present novel adaptations of two major shift-reduce dependency parsing algorithms to character-level parsing. Experimental results on the Chinese Treebank demonstrate improved performances over word-based parsing methods.
Our first impression of others is highly influenced by their facial appearance. However, the perception and evaluation of faces is not only guided by internal features such as facial expressions, but also highly dependent on contextual information such as secondhand information (verbal descriptions) about the target person. To investigate the time course of contextual influences on cortical face processing, event-related brain potentials were investigated in response to neutral faces, which were preceded by brief verbal descriptions containing cues of affective valence (negative, neutral, positive) and self-reference (self-related vs. other-related). ERP analysis demonstrated that early and late stages of face processing are enhanced by negative and positive as well as self-relevant descriptions, although faces per se did not differ perceptually. Affective ratings of the faces confirmed these findings. Altogether, these results demonstrate for the first time both on an electrocortical and behavioral level how contextual information modifies early visual perception in a top-down manner.
Methylphenidate mainly enhances dopamine neurotransmission whereas 3,4-methylenedioxymethamphetamine (MDMA, "ecstasy") mainly enhances serotonin neurotransmission. However, both drugs also induce a weaker increase of cerebral noradrenaline exerting sympathomimetic properties. Dopaminergic psychostimulants are reported to increase sexual drive, while serotonergic drugs typically impair sexual arousal and functions. Additionally, serotonin has also been shown to modulate cognitive perception of romantic relationships. Whether methylphenidate or MDMA alter sexual arousal or cognitive appraisal of intimate relationships is not known. Thus, we evaluated effects of methylphenidate (40 mg) and MDMA (75 mg) on subjective sexual arousal by viewing erotic pictures and on perception of romantic relationships of unknown couples in a double-blind, randomized, placebo-controlled, crossover study in 30 healthy adults. Methylphenidate, but not MDMA, increased ratings of sexual arousal for explicit sexual stimuli. The participants also sought to increase the presentation time of implicit sexual stimuli by button press after methylphenidate treatment compared with placebo. Plasma levels of testosterone, estrogen, and progesterone were not associated with sexual arousal ratings. Neither MDMA nor methylphenidate altered appraisal of romantic relationships of others. The findings indicate that pharmacological stimulation of dopaminergic but not of serotonergic neurotransmission enhances sexual drive. Whether sexual perception is altered in subjects misusing methylphenidate e.g., for cognitive enhancement or as treatment for attention deficit hyperactivity disorder is of high interest and warrants further investigation.
This is the first attempt at characterizing reading difficulty in Hindi using naturally occurring sentences. We created the Potsdam-Allahabad Hindi Eyetracking Corpus by recording eye-movement data from 30 participants at the University of Allahabad, India. The target stimuli were 153 sentences selected from the beta version of the Hindi-Urdu treebank. We find that word- or low-level predictors (syllable length, unigram and bigram frequency) affect first-pass reading times, regression path duration, total reading time, and outgoing saccade length. An increase in syllable length results in longer fixations, and an increase in word unigram and bigram frequency leads to shorter fixations. Longer syllable length and higher frequency lead to longer outgoing saccades. We also find that two predictors of sentence comprehension difficulty, integration and storage cost, have an effect on reading difficulty. Integration cost (Gibson, 2000) was approximated by calculating the distance (in words) between a dependent and head; and storage cost (Gibson, 2000), which measures difficulty of maintaining predictions, was estimated by counting the number of predicted heads at each point in the sentence. We find that integration cost mainly affects outgoing saccade length, and storage cost affects total reading times and outgoing saccade length. Thus, word-level predictors have an effect in both early and late measures of reading time, while predictors of sentence comprehension difficulty tend to affect later measures. This is, to our knowledge, the first demonstration using eye-tracking that both integration and storage cost influence reading difficulty.
This paper proposes a simple yet effective framework of soft cross-lingual syntax projection to transfer syntactic structures from source language to target language using monolingual treebanks and large-scale bilingual parallel text. Here, soft means that we only project reliable dependencies to compose high-quality target structures. The projected instances are then used as additional training data to improve the performance of supervised parsers. The major issues for this idea are 1) errors from the source-language parser and unsupervised word aligner; 2) intrinsic syntactic non-isomorphism between languages; 3) incomplete parse trees after projection. To handle the first two issues, we propose to use a probabilistic dependency parser trained on the target-language treebank, and prune out unlikely projected dependencies that have low marginal probabilities. To make use of the incomplete projected syntactic structures, we adopt a new learning technique based on ambiguous labelings. For a word that has no head words after projection, we enrich the projected structure with all other words as its candidate heads as long as the newly-added dependency does not cross any projected dependencies. In this way, the syntactic structure of a sentence becomes a parse forest (ambiguous labels) instead of a single parse tree. During training, the objective is to maximize the mixed likelihood of manually labeled instances and projected instances with ambiguous labelings. Experimental results on benchmark data show that our method significantly outperforms a strong baseline supervised parser and previous syntax projection methods. 1
This study examines the methodology of global foreign accent ratings in studies on L2 speech production. In three experiments, we test how variation in raters, range within speech samples, as well as instructions and procedures affects ratings of accent in predominantly monolingual speakers of German, non-native speakers of German, as well as long-term emigrants from Germany, that is, L1 attriters. The findings show that rater differences do not result in systematic changes in rating patterns. In contrast, range effects and effects of familiarity with accented speech lead to shifts in absolute and relative ratings. Including more strongly foreign-accented samples leads to lower judgements for the entire group of L2 speakers compared to natives. Similarly, lower familiarity with foreign accent results in more variable and more strongly foreign-accented judgements. We discuss the implications for research on L2 pronunciation as well as for the interpretation of nativeness in L2 studies and language testing more generally.
This chapter investigates problems that novice writers, especially learners of English, have in acquiring lexical features of written discourse. It describes features of an interactive writing kit (WordPilot) which assists them to access authentic text, then to transfer patterns they explore there to their own writing. Relevant text corpora and specific lexical databases are made available during the writing process through the mediation of this integrated electronic reference, which incorporates several applications, including a concordancer, dictionary and thesaurus. This learning and production device has been implemented for Cantonese-speaking students in Hong Kong for whom English is a Foreign Language, but the general principles have relevance to writers at any point on the continuum of language proficiency. The chapter also illustrates problems in the presentation of lexical patterns to novice writers of limited English proficiency. As a gateway to text, it provides novice writers opportunities to explore language features, and it assists academic gate keepers in initiating learners into discourse.
In this paper, we report the obtained results of two constituency parsers trained with BulTreeBank, an HPSG-based treebank for Bulgarian. To reduce the data sparsity problem, we propose using the Brown word clustering to do an off-line clustering and map the words in the treebank to create a class-based treebank. e observations show that when the classes outnumber the POS tags, the results are beer. Since this approach adds on another dimension of abstraction (in comparison to the lemma), its coarse-grained representation can be used further for training statistical parsers.
A phrase dependency treebank(PDT)integrating phrase structure grammar and dependency grammar is proposed and elaborated to cater for translation studies.The construction of DUT Parallel Chinese-English PDT(DUT-CEPDT)is reported.PDT favors flat structures and the dependency is based on semantics rather than syntactic functions,which differs from the mainstream dependency analysis that favors binary branching.The raw texts of DUT-CEPDT are Chinese government work reports and White Papers and their official English translation.First of all,after word segmentation and part of speech(POS)tagging,Chinese PDT and English PDT are constructed manually with the aid of LingTreeConstructor,a tool tailored for linguists.Then,node alignment, which covers translation alignments of words,phrases,clauses up to the whole passage,is proposed instead of traditional word or sentence alignment to provide more translation knowledge.Lastly, semantic roles based on the FrameNet are labeled simultaneously on the aligned nodes of the English and Chinese trees.DUT-CEPDT can serve as a resource and standard of the training and assessment of both human translators and machine translation systems.
We present novel computational experiments using William Labov’s theory of narrative analysis. We describe his six elements of narrative structure and construct a new corpus based on his most recent work on narrative. Using this corpus, we explore the correspondence between Labovs elements of narrative structure and the implicit discourse relations of the Penn Discourse Treebank, and we construct a mapping between the elements of narrative structure and the discourse relation classes of the PDTB. We present first experiments on detecting Complicating Actions, the most common of the elements of narrative structure, achieving an f-score of 71.55. We compare the contributions of features derived from narrative analysis, such as the length of clauses and the tenses of main verbs, with those of features drawn from work on detecting implicit discourse relations. Finally, we suggest directions for future research on narrative structure, such as applications in assessing text quality and in narrative generation.
BACKGROUND: The standard clinical acquisition for left ventricular functional parameter analysis with cardiovascular magnetic resonance (CMR) uses a multi-breathhold multi-slice segmented balanced SSFP sequence. Performing multiple long breathholds in quick succession for ventricular coverage in the short-axis orientation can lead to fatigue and is challenging in patients with severe cardiac or respiratory disorders. This study combines the encoding efficiency of a six-fold undersampled 3D stack of spirals balanced SSFP sequence with 3D through-time spiral GRAPPA parallel imaging reconstruction. This 3D spiral method requires only one breathhold to collect the dynamic data. METHODS: Ten healthy volunteers were recruited for imaging at 3 T. The 3D spiral technique was compared against 2D imaging in terms of systolic left ventricular functional parameter values (Bland-Altman plots), total scan time (Welch's t-test) and qualitative image rating scores (Wilcoxon signed-rank test). RESULTS: Systolic left ventricular functional values were not significantly different (i.e. 3D-2D) between the methods. The 95% confidence interval for ejection fraction was -0.1 ± 1.6% (mean ± 1.96*SD). The total scan time for the 3D spiral technique was 48 s, which included one breathhold with an average duration of 14 s for the dynamic scan, plus 34 s to collect the calibration data under free-breathing conditions. The 2D method required an average of 5 min 40s for the same coverage of the left ventricle. The difference between 3D and 2D image rating scores was significantly different from zero (Wilcoxon signed-rank test, p < 0.05); however, the scores were at least 3 (i.e. average) or higher for 3D spiral imaging. CONCLUSION: The 3D through-time spiral GRAPPA method demonstrated equivalent systolic left ventricular functional parameter values, required significantly less total scan time and yielded acceptable image quality with respect to the 2D segmented multi-breathhold standard in this study. Moreover, the 3D spiral technique used just one breathhold for dynamic imaging, which is anticipated to reduce patient fatigue as part of the complete cardiac examination in future studies that include patients.
Discourse relations bind smaller linguistic units into coherent texts. However, automatically identifying discourse relations is difficult, because it requires understanding the semantics of the linked arguments. A more subtle challenge is that it is not enough to represent the meaning of each argument of a discourse relation, because the relation may depend on links between lower-level components, such as entity mentions. Our solution computes distributional meaning representations by composition up the syntactic parse tree. A key difference from previous work on compositional distributional semantics is that we also compute representations for entity mentions, using a novel downward compositional pass. Discourse relations are predicted from the distributional representations of the arguments, and also of their coreferent entity mentions. The resulting system obtains substantial improvements over the previous state-of-the-art in predicting implicit discourse relations in the Penn Discourse Treebank.
This work improves a novel Service Selection Method for the development of Service-Oriented Applications in the context of the Service-Oriented Computing (SOC) paradigm. We have defined a Semantic-Structural Scheme to assess Web Services on Interface Compatibility exploring the available information from WSDL documents. The structural information involves data types from return, parameters and exceptions. The semantic information concerns identifiers from parameters and operation names. The lexical database WordNet is used as a semantic basis. Two appraisal values were defined: compatibility gap and adaptability gap. The former is centered on functional aspects. The latter explains the adaptation effort to a successful integration. We validated those appraisals values through different experiments with a data-set of 465 real-life Web Services and measured the results using three metrics from the Information Retrieval field.
This paper gives an overview of the latest developments in computational syntactic analysis of Estonian. We present Estonian Dependency Treebank, an ongoing corpus annotation project. Although the treebank construction is still under way, we have used it for training MaltParser and experimenting with combining MaltParser with a rule-based Constraint Grammar parser for Estonian. MaltParser achieves unlabeled attachment score (UAS; correct links to head node) of 83.4% and label accuracy (LA) of 88.6%. Labeled attachment score (LAS) was 80.3%.
Abstract Statistical parsers often require careful parameter tuning and feature selection. This is a nontrivial task for application developers who are not interested in parsing for its own sake, and it can be time-consuming even for experienced researchers. In this paper we present MaltOptimizer, a tool developed to automatically explore parameters and features for MaltParser, a transition-based dependency parsing system that can be used to train parser's given treebank data. MaltParser provides a wide range of parameters for optimization, including nine different parsing algorithms, an expressive feature specification language that can be used to define arbitrarily rich feature models, and two machine learning libraries, each with their own parameters. MaltOptimizer is an interactive system that performs parser optimization in three stages. First, it performs an analysis of the training set in order to select a suitable starting point for optimization. Second, it selects the best parsing algorithm and tunes the parameters of this algorithm. Finally, it performs feature selection and tunes machine learning parameters. Experiments on a wide range of data sets show that MaltOptimizer quickly produces models that consistently outperform default settings and often approach the accuracy achieved through careful manual optimization.