Refine
Year of publication
- 2009 (23) (remove)
Document Type
- Conference Proceeding (23) (remove)
Has Fulltext
- yes (23)
Is part of the Bibliography
- no (23)
Keywords
- Computerlinguistik (4)
- Deutsch (4)
- Korpus <Linguistik> (4)
- Automatische Sprachanalyse (3)
- Mehrsprachigkeit (3)
- Natürliche Sprache (3)
- Algorithmus (2)
- Bildung (2)
- Datensatz (2)
- Polarität (2)
Publicationstate
- Veröffentlichungsversion (11)
- Zweitveröffentlichung (8)
- Postprint (4)
Reviewstate
- Peer-Review (11)
- (Verlags)-Lektorat (5)
- Review-Status-unbekannt (1)
Publisher
- AAAI Press (1)
- AKS-Verlag (1)
- Acta Press (1)
- Acta Universitatis Upsaliensis (1)
- Association for Computational Linguistics (1)
- Association for Computing Machinery (1)
- Berlin-Brandenburgische Akademie der Wissenschaften; Zentrum Sprache (1)
- CSLI Publications (1)
- Department of Linguistics, University of Cambridge (1)
- EDUCatt (1)
In opinion mining, there has been only very little work investigating semi-supervised machine learning on document-level polarity classification. We show that semi-supervised learning performs significantly better than supervised learning when only few labelled data are available. Semi-supervised polarity classifiers rely on a predictive feature set. (Semi-)Manually built polarity lexicons are one option but they are expensive to obtain and do not necessarily work in an unknown domain. We show that extracting frequently occurring adjectives & adverbs of an unlabeled set of in-domain documents is an inexpensive alternative which works equally well throughout different domains.
Though polarity classification has been extensively explored at document level, there has been little work investigating feature design at sentence level. Due to the small number of words within a sentence, polarity classification at sentence level differs substantially from document-level classification in that resulting bag-of-words feature vectors tend to be very sparse resulting in a lower classification accuracy.
In this paper, we show that performance can be improved by adding features specifically designed for sentence-level polarity classification. We consider both explicit polarity information and various linguistic features. A great proportion of the improvement that can be obtained by using polarity information can also be achieved by using a set of simple domain-independent linguistic features.
Generative lexicalized parsing models, which are the mainstay for probabilistic parsing of English, do not perform as well when applied to languages with different language-specific properties such as free(r) word order or rich morphology. For German and other non-English languages, linguistically motivated complex treebank transformations have been shown to improve performance within the framework of PCFG parsing, while generative lexicalized models do not seem to be as easily adaptable to these languages. In this paper, we show a practical way to use grammatical functions as first-class citizens in a discriminative model that allows to extend annotated treebank grammars with rich feature sets without having to suffer from sparse data problems. We demonstrate the flexibility of the approach by integrating unsupervised PP attachment and POS-based word clusters into the parser.
From Proof Texts to Logic. Discourse Representation Structures for Proof Texts in Mathematics
(2009)
We present an extension to Discourse Representation Theory that can be used to analyze mathematical texts written in the commonly used semi-formal language of mathematics (or at least a subset of it). Moreover, we describe an algorithm that can be used to check the resulting Proof Representation Structures for their logical validity and adequacy as a proof.
TEI Feature Structures as a Representation Format for Multiple Annotation and Generic XML Documents
(2009)
Feature structures are mathematical entities (rooted labeled directed acyclic graphs) that can be represented as graph displays, attribute value matrices or as XML adhering to the constraints of a specialized TEI tag set. We demonstrate that this latter ISO-standardized format can be used as an integrative storage and exchange format for sets of multiple annotation XML documents. This specific domain of application is rooted in the approach of multiple annotations, which marks a possible solution for XML-compliant markup in scenarios with conflicting annotation hierarchies. A more extreme proposal consists in the possible use as a meta-representation format for generic XML documents. For both scenarios our strategy concerning pertinent feature structure representations is grounded on the XDM (XQuery 1.0 and XPath 2.0 Data Model). The ubiquitous hierarchical and sequential relationships within XML documents are represented by specific features that take ordered list values. The mapping to the TEI feature structure format has been implemented in the form of an XSLT 2.0 stylesheet. It can be characterized as exploiting aspects of both the push and pull processing paradigm as appropriate. An indexing mechanism is provided with regard to the multiple annotation documents scenario. Hence, implicit links concerning identical primary data are made explicit in the result format. In comparison to alternative representations, the TEI-based format does well in many respects, since it is both integrative and well-formed XML. However, the result documents tend to grow very large depending on the size of the input documents and their respective markup structure. This may also be considered as a downside regarding the proposed use for generic XML documents. On the positive side, it may be possible to achieve a hookup to methods and applications that have been developed for feature structure representations in the fields of (computational) linguistics and knowledge representation.
The paper discusses particular logical consistency conditions satisfied by German proposition-embedding predicates which determine the question type (external and internal whether-form as well as exhaustive and non-exhaustive wh-form), the correlate type (es- or da-correlate) as well as the impact of the correlate on the respective consistency condition. It will turn out that some consistency conditions also determine the embedding of verb second and subject-control.
This paper describes a new approach to improve the analysis and categorization of web documents using statistical methods for template based clustering as well as semantical analysis based on terminological ontologies. A domain-specific environment serves for prove of concept. In order to demonstrate the widespread practical benefit of our approach, we outline a combined mathematical and semantical framework for information retrieval on internet resources.
Spoken language corpora— as used in conversation analytic research, language acquisition studies and dialectology— pose a number of challenges that are rarely addressed by corpus linguistic methodology and technology. This paper starts by giving an overview of the most important methodological issues distinguishing spoken language corpus workfrom the work with written data. It then shows what technological challenges these methodological issues entail and demonstrates how they are dealt with in the architecture and tools of the EXMARaLDA system.
We present data-driven methods for the acquisition of LFG resources from two German treebanks. We discuss problems specific to semi-free word order languages as well as problems arising from the data structures determined by the design of the different treebanks. We compare two ways of encoding semi-free word order, as done in the two German treebanks, and argue that the design of the TiGer treebank is more adequate for the acquisition of LFG resources. Furthermore, we describe an architecture for LFG grammar acquisition for German, based on the two German treebanks, and compare our results with a hand-crafted German LFG grammar.
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.