Refine
Year of publication
Document Type
- Conference Proceeding (26) (remove)
Language
- English (26)
Has Fulltext
- yes (26)
Is part of the Bibliography
- no (26) (remove)
Keywords
- Korpus <Linguistik> (15)
- Annotation (12)
- Syntaktische Analyse (9)
- Automatische Sprachanalyse (8)
- Deutsch (5)
- Frame-Semantik (3)
- Gesprochene Sprache (3)
- Lexikalisch funktionale Grammatik (2)
- Active Learning (1)
- Active learning (1)
Publicationstate
Reviewstate
- Peer-Review (10)
- (Verlags)-Lektorat (9)
Publisher
TePaCoC - A Testsuite for Testing Parser Performance on Complex German Grammatical Constructions
(2009)
The Stuttgart-Tübingen 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.
We present MaJo, a toolkit for supervised Word Sense Disambiguation (WSD), with an interface for Active Learning. Our toolkit combines a flexible plugin architecture which can easily be extended, with a graphical user interface which guides the user through the learning process. MaJo integrates off-the-shelf NLP tools like POS taggers, treebank-trained statistical parsers, as well as linguistic resources like WordNet and GermaNet. It enables the user to systematically explore the benefit gained from different feature types for WSD. In addition, MaJo provides an Active Learning environment, where the
system presents carefully selected instances to a human oracle. The toolkit supports manual annotation of the selected instances and re-trains the system on the extended data set. MaJo also provides the means to evaluate the performance of the system against a gold standard. We illustrate the usefulness of our system by learning the frames (word senses) for three verbs from the SALSA corpus, a version of the TiGer treebank with an additional layer of frame-semantic annotation. We show how MaJo can be used to tune the feature set for specific target words and so improve performance for these targets. We also show that syntactic features, when carefully tuned to the target word, can lead to a substantial increase in performance.
The annotation of parts of speech (POS) in linguistically annotated corpora is a fundamental annotation layer which provides the basis for further syntactic analyses, and many NLP tools rely on POS information as input. However, most POS annotation schemes have been developed with written (newspaper) text in mind and thus do not carry over well to text from other domains and genres. Recent discussions have concentrated on the shortcomings of present POS annotation schemes with regard to their applicability to data from domains other than newspaper text.
In the NLP literature, adapting a parser to new text with properties different from the training data is commonly referred to as domain adaptation. In practice, however, the differences between texts from different sources often reflect a mixture of domain and genre properties, and it is by no means clear what impact each of those has on statistical parsing. In this paper, we investigate how differences between articles in a newspaper corpus relate to the concepts of genre and domain and how they influence parsing performance of a transition-based dependency parser. We do this by applying various similarity measures for data point selection and testing their adequacy for creating genre-aware parsing models.
This paper presents Release 2.0 of the SALSA corpus, a German resource for lexical semantics. The new corpus release provides new annotations for German nouns, complementing the existing annotations of German verbs in Release 1.0. The corpus now includes around 24,000 sentences with more than 36,000 annotated instances. It was designed with an eye towards NLP applications such as semantic role labeling but will also be a useful resource for linguistic studies in lexical semantics.
Current work on sentiment analysis is characterized by approaches with a pragmatic focus, which use shallow techniques in the interest of robustness but often rely on ad-hoc creation of data sets and methods. We argue that progress towards deep analysis depends on a) enriching shallow representations with linguistically motivated, rich information, and b) focussing different branches of research and combining ressources to create synergies with related work in NLP. In the paper, we propose SentiFrameNet, an extension to FrameNet, as a novel representation for sentiment analysis that is tailored to these aims.
This paper presents an extension to the Stuttgart-Tübingen TagSet, the standard part-of-speech tag set for German, for the annotation of spoken language. The additional tags deal with hesitations, backchannel signals, interruptions, onomatopoeia and uninterpretable material. They allow one to capture phenomena specific to spoken language while, at the same time, preserving inter-operability with already existing corpora of written language.
In the paper we investigate the impact of data size on a Word Sense Disambiguation task (WSD). We question the assumption that the knowledge acquisition bottleneck, which is known as one of the major challenges for WSD, can be solved by simply obtaining more and more training data. Our case study on 1,000 manually annotated instances of the German verb drohen (threaten) shows that the best performance is not obtained when training on the full data set, but by carefully selecting new training instances with regard to their informativeness for the learning process (Active Learning). We present a thorough evaluation of the impact of different sampling methods on the data sets and propose an improved method for uncertainty sampling which dynamically adapts the selection of new instances to the learning progress of the classifier, resulting in more robust results during the initial stages of learning. A qualitative error analysis identifies problems for automatic WSD and discusses the reasons for the great gap in performance between human annotators and our automatic WSD system.
Active Learning (AL) has been proposed as a technique to reduce the amount of annotated data needed in the context of supervised classification. While various simulation studies for a number of NLP tasks have shown that AL works well on goldstandard data, there is some doubt whether the approach can be successful when applied to noisy, real-world data sets. This paper presents a thorough evaluation of the impact of annotation noise on AL and shows that systematic noise resulting from biased coder decisions can seriously harm the AL process. We present a method to filter out inconsistent annotations during AL and show that this makes AL far more robust when applied to noisy data.
Problems for parsing morphologically rich languages are, amongst others, caused by the higher variability in structure due to less rigid word order constraints and by the higher number of different lexical forms. Both properties can result in sparse data problems for statistical parsing. We present a simple approach for addressing these issues. Our approach makes use of self-training on instances selected with regard to their similarity to the annotated data. Our similarity measure is based on the perplexity of part-of-speech trigrams of new instances measured against the annotated training data. Preliminary results show that our method outperforms a self-training setting where instances are simply selected by order of occurrence in the corpus and argue that selftraining is a cheap and effective method for improving parsing accuracy for morphologically rich languages.
This paper contributes to the discussion on best practices for the syntactic analysis of non-canonical language, focusing on Twitter microtext. We present an annotation experiment where we test an existing POS tagset, the Stuttgart-Tübingen Tagset (STTS), with respect to its applicability for annotating new text from the social media, in particular from Twitter microblogs. We discuss different tagset extensions proposed in the literature and test our extended tagset on a set of 506 tweets (7.418 tokens) where we achieve an inter-annotator agreement for two human annotators in the range of 92.7 to 94.4 (k). Our error analysis shows that especially the annotation of Twitterspecific phenomena such as hashtags and at-mentions causes disagreements between the human annotators. Following up on this, we provide a discussion of the different uses of the @- and #-marker in Twitter and argue against analysing both on the POS level by means of an at-mention or hashtag label. Instead, we sketch a syntactic analysis which describes these phenomena by means of syntactic categories and grammatical functions.
For languages with (semi-) free word order (such as German), labelling grammatical functions on top of phrase-structural constituent analyses is crucial for making them interpretable. Unfortunately, most statistical classifiers consider only local information for function labelling and fail to capture important restrictions on the distribution of core argument functions such as subject, object etc., namely that there is at most one subject (etc.) per clause. We augment a statistical classifier with an integer linear program imposing hard linguistic constraints on the solution space output by the classifier, capturing global distributional restrictions. We show that this improves labelling quality, in particular for argument grammatical functions, in an intrinsic evaluation, and, importantly, grammar coverage for treebankbased (Lexical-Functional) grammar acquisition and parsing, in an extrinsic evaluation.
Recent work on error detection has shown that the quality of manually annotated corpora can be substantially improved by applying consistency checks to the data and automatically identifying incorrectly labelled instances. These methods, however, can not be used for automatically annotated corpora where errors are systematic and cannot easily be identified by looking at the variance in the data. This paper targets the detection of POS errors in automatically annotated corpora, so-called silver standards, showing that by combining different measures sensitive to annotation quality we can identify a large part of the errors and obtain a substantial increase in accuracy.
This paper presents an annotation scheme for English modal verbs together with sense-annotated data from the news domain. We describe our annotation scheme and discuss problematic cases for modality annotation based on the inter-annotator agreement during the annotation. Furthermore, we present experiments on automatic sense tagging, showing that our annotations do provide a valuable training resource for NLP systems.
This paper is a contribution to the ongoing discussion on treebank annotation schemes and their impact on PCFG parsing results. We provide a thorough comparison of two German treebanks: the TIGER treebank and the TüBa-D/Z. We use simple statistics on sentence length and vocabulary size, and more refined methods such as perplexity and its correlation with PCFG parsing results, as well as a Principal Components Analysis. Finally we present a qualitative evaluation of a set of 100 sentences from the TüBa- D/Z, manually annotated in the TIGER as well as in the TüBa-D/Z annotation scheme, and show that even the existence of a parallel subcorpus does not support a straightforward and easy comparison of both annotation schemes.
This paper presents a thorough examination of the validity of three evaluation measures on parser output. We assess parser performance of an unlexicalised probabilistic parser trained on two German treebanks with different annotation schemes and evaluate parsing results using the PARSEVAL metric, the Leaf-Ancestor metric and a dependency-based evaluation. We reject the claim that the TüBa-D/Z annotation scheme is more adequate then the TIGER scheme for PCFG parsing and show that PARSEVAL should not be used to compare parser performance for parsers trained on treebanks with different annotation schemes. An analysis of specific error types indicates that the dependency-based evaluation is most appropriate to reflect parse quality.
Recent studies focussed on the question whether less-configurational languages like German are harder to parse than English, or whether the lower parsing scores are an artefact of treebank encoding schemes and data structures, as claimed by Kübler et al. (2006). This claim is based on the assumption that PARSEVAL metrics fully reflect parse quality across treebank encoding schemes. In this paper we present new experiments to test this claim. We use the PARSEVAL metric, the Leaf-Ancestor metric as well as a dependency-based evaluation, and present novel approaches measuring the effect of controlled error insertion on treebank trees and parser output. We also provide extensive past-parsing crosstreebank conversion. The results of the experiments show that, contrary to Kübler et al. (2006), the question whether or not German is harder to parse than English remains undecided.
This paper discusses the behaviour of German particle verbs formed by two-way prepositions in combination with pleonastic PPs including the verb particle as a preposition. These particle verbs have a characteristic feature: some of them license directional prepositional phrases in the accusative, some only allow for locative PPs in the dative, and some particle verbs can occur with PPs in the accusative and in the dative. Directional particle verbs together with directional PPs present an additional problem: the particle and the preposition in the PP seem to provide redundant information. The paper gives an overview of the semantic verb classes influencing this phenomenon, based on corpus data, and explains the underlying reasons for the behaviour of the particle verbs. We also show how the restrictions on particle verbs and pleonastic PPs can be expressed in a grammar theory like Lexical Functional Grammar (LFG).