Refine
Year of publication
Document Type
- Conference Proceeding (52)
- Part of a Book (11)
- Article (9)
- Book (1)
- Doctoral Thesis (1)
- Working Paper (1)
Language
- English (75) (remove)
Keywords
- Automatische Sprachanalyse (75) (remove)
Publicationstate
- Veröffentlichungsversion (54)
- Postprint (11)
- Zweitveröffentlichung (6)
Reviewstate
Publisher
- Association for Computational Linguistics (10)
- German Society for Computational Linguistics & Language Technology und Friedrich-Alexander-Universität Erlangen-Nürnberg (5)
- Springer (5)
- European Language Resources Association (ELRA) (4)
- LiU Electronic Press (4)
- The Association for Computational Linguistics (4)
- Association for Computing Machinery (2)
- European Language Resources Association (2)
- European language resources association (ELRA) (2)
- Gesellschaft für Sprachtechnologie and Computerlinguistik e.V. (2)
Authors like Fillmore 1986 and Goldberg 2006 have made a strong case for regarding argument omission in English as a lexical and construction-based affordance rather than one based on general semantico-pragmatic constraints. They do not, however, address the question of how grammatical restrictions on null complementation might interact with broader narrative conventions, in particular those of genre. In this paper, we attempt to remedy this oversight by presenting a comprehensive overview of genre-based argument omissions and offering a construction-based analysis of genre-based omission conventions. We consider five genre-based omission types: instructional imperatives (Culy 1996, Bender 1999), labelese, diary style (Haegeman 1990), match reports (Ruppenhofer 2004) and quotative clauses. We show that these omission types share important traits; all, for example, have anaphoric rather than indefinite construals. We also show, however, that the omission types differ from each other in idiosyncratic ways. We then address several interrelated representational problems posed by the grammatical treatment of genre-based omissions. For example, the constructions that represent genre-based omission conventions must interact with the lexical entries of verbs, many of which do not generally permit omitted arguments. Accordingly, we offer constructional analyses of genre-based omissions that allow constructions to override lexical valence constraints.
There have been several attempts to annotate communicative functions to utterances of verbal feedback in English previously. Here, we suggest an annotation scheme for verbal and non-verbal feedback utterances in French including the categories base, attitude, previous and visual. The data comprises conversations, maptasks and negotiations from which we extracted ca. 13,000 candidate feedback utterances and gestures. 12 students were recruited for the annotation campaign of ca. 9,500 instances. Each instance was annotated by between 2 and 7 raters. The evaluation of the annotation agreement resulted in an average best-pair kappa of 0.6. While the base category with the values acknowledgement, evaluation, answer, elicit and other achieves good agreement, this is not the case for the other main categories. The data sets, which also include automatic extractions of lexical, positional and acoustic features, are freely available and will further be used for machine learning classification experiments to analyse the form-function relationship of feedback.
We present a descriptive analysis on the two datasets from the shared task on Source, Subjective Expression and Target Extraction from Political Speeches (STEPS), the only existing German dataset for opinion role extraction of its size. Our analysis discusses the individual properties of the three components, subjective expressions, sources and targets and their relations towards each other. Our observations should help practitioners and researchers when building a system to extract opinion roles from German data.
A Supervised learning approach for the extraction of opinion sources and targets from German text
(2019)
We present the first systematic supervised learning approach for the extraction of opinion sources and targets on German language data. A wide choice of different features is presented, particularly syntactic features and generalization features. We point out specific differences between opinion sources and targets. Moreover, we explain why implicit sources can be extracted even with fairly generic features. In order to ensure comparability our classifier is trained and tested on the dataset of the STEPS shared task.
This contribution presents an XML Schema for annotating a high level narratological category: speech, thought and writing representation (ST&WR). It focusses on two aspects: Firstly, the original Schema is presented as an example for the challenge to encode a narrative feature in a structured and flexible way and secondly, ways of adapting this Schema to TEI are considered, in Order to make it usable for other, TEI-based projects.
Distributional models of word use constitute an indispensable tool in corpus based lexicological research for discovering paradigmatic relations and syntagmatic patterns (Belica et al. 2010). Recently, word embeddings (Mikolov et al. 2013) have revived the field by allowing to construct and analyze distributional models on very large corpora. This is accomplished by reducing the very high dimensionality of word cooccurrence contexts, the size of the vocabulary, to few dimensions, such as 100-200. However, word use and meaning can vary widely along dimensions such as domain, register, and time, and word embeddings tend to represent only the most prevalent meaning. In this paper we thus construct domain specific word embeddings to allow for systematically analyzing variations in word use. Moreover, we also demonstrate how to reconstruct domain specific co-occurrence contexts from the dense word embeddings.
This paper discusses computational linguistic methods for the semi-automatic analysis of modality interdependencies (the combination of complex resources such as speaking, writing, and visualizing; MID) in professional crosssituational interaction settings. The overall purpose of the approach is to develop models, methods, and a framework for the description and analysis of MID forms and functions. The paper describes work in progress—the development of an annotation framework that allows annotating different data and file formats at various levels, to relate annotation levels and entries independently of the given file format, and to visualize patterns.
In this paper we present the results of an automatic classification of Russian texts into three levels of difficulty. Our aim is to build a study corpus of Russian, in which a L2 student is able to select texts of a desired complexity. We are building on a pilot study, in which we classified Russian texts into two levels of difficulty. In the current paper, we apply the classification to an extended corpus of 577 labelled texts. The best-performing combination of features achieves an accuracy of 0,74 within at most one level difference.
Beyond the stars: exploiting free-text user reviews to improve the accuracy of movie recommendations
(2009)
In this paper we show that the extraction of opinions from free-text reviews can improve the accuracy of movie recommendations. We present three approaches to extract movie aspects as opinion targets and use them as features for the collaborative filtering. Each of these approaches requires different amounts of manual interaction. We collected a data set of reviews with corresponding ordinal (star) ratings of several thousand movies to evaluate the different features for the collaborative filtering. We employ a state-of-the-art collaborative filtering engine for the recommendations during our evaluation and compare the performance with and without using the features representing user preferences mined from the free-text reviews provided by the users. The opinion mining based features perform significantly better than the baseline, which is based on star ratings and genre information only.
This paper presents C-WEP, the Collection of Writing Errors by Professionals Writers of German. It currently consists of 245 sentences with grammatical errors. All sentences are taken from published texts. All authors are professional writers with high skill levels with respect to German, the genres, and the topics. The purpose of this collection is to provide seeds for more sophisticated writing support tools as only a very small proportion of those errors can be detected by state-of-the-art checkers. C-WEP is annotated on various levels and freely available.
Catching the common cause: extraction and annotation of causal relations and their participants
(2017)
In this paper, we present a simple, yet effective method for the automatic identification and extraction of causal relations from text, based on a large English-German parallel corpus. The goal of this effort is to create a lexical resource for German causal relations. The resource will consist of a lexicon that describes constructions that trigger causality as well as the participants of the causal event, and will be augmented by a corpus with annotated instances for each entry, that can be used as training data to develop a system for automatic classification of causal relations. Focusing on verbs, our method harvested a set of 100 different lexical triggers of causality, including support verb constructions. At the moment, our corpus includes over 1,000 annotated instances. The lexicon and the annotated data will be made available to the research community.
The FrameNet lexical database yields information about collocations and multiword expressions in various ways. In some cases phrasal units have been entered from the start as lexical entries (write down). In other cases headword + preposition pairs can be recognized as special collocations Where the preposition in question is a necessary and lexically specified marker of an argument of the headword + fond of, hostile to). Nominal compounds are annotated with respect to noun or (pertinative) adjective modifiers, some of which are analyzable but also entrenched (wheel chair, fiscal year). Nouns that name aggregates, portions, types, etc., sometimes hold lexically specified relations to their dependents (flock of geese). And event nouns frequently Select the support verbs which permit them to enter into predications (file an objection, enter a plea). A subproject aims at extracting, as structured clusters of lexical items, the minimal semantically central kernel dependency graphs from the set of annotations. Such research will yield not only commonplace groupings (eat: dog, bone) but will also yield hitherto unnoticed collocations within such graphs (answer: you, door) where certain dependency links within them are idiomatic or otherwise lexically special, here answer > door. Collocational information can also be retrieved by various types of queries within our MySQL search tool
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.
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.
The automatic recognition of idioms poses a challenging problem for NLP applications. Whereas native speakers can intuitively handle multiword expressions whose compositional meanings are hard to trace back to individual word semantics, there is still ample scope for improvement regarding computational approaches. We assume that idiomatic constructions can be characterized by gradual intensities of semantic non-compositionality, formal fixedness, and unusual usage context, and introduce a number of measures for these characteristics, comprising count-based and predictive collocation measures together with measures of context (un)similarity. We evaluate our approach on a manually labelled gold standard, derived from a corpus of German pop lyrics. To this end, we apply a Random Forest classifier to analyze the individual contribution of features for automatically detecting idioms, and study the trade-off between recall and precision. Finally, we evaluate the classifier on an independent dataset of idioms extracted from a list of Wikipedia idioms, achieving state-of-the art accuracy.
In this paper, we present our work-inprogress to automatically identify free indirect representation (FI), a type of thought representation used in literary texts. With a deep learning approach using contextual string embeddings, we achieve f1 scores between 0.45 and 0.5 (sentence-based evaluation for the FI category) on two very different German corpora, a clear improvement on earlier attempts for this task. We show how consistently marked direct speech can help in this task. In our evaluation, we also consider human inter-annotator scores and thus address measures of certainty for this difficult phenomenon.
We examine the new task of detecting derogatory compounds (e.g. curry muncher). Derogatory compounds are much more difficult to detect than derogatory unigrams (e.g. idiot) since they are more sparsely represented in lexical resources previously found effective for this task (e.g. Wiktionary). We propose an unsupervised classification approach that incorporates linguistic properties of compounds. It mostly depends on a simple distributional representation. We compare our approach against previously established methods proposed for extracting derogatory unigrams.
Automatic division of spoken language transcripts into sentence-like units is a challenging problem, caused by disfluencies, ungrammatical structures and the lack of punctuation. We present experiments on dividing up German spoken dialogues where we investigate the impact of task setup and data representation, encoding of context information as well as different model architectures for this task.
We discuss the impact of data bias on abusive language detection. We show that classification scores on popular datasets reported in previous work are much lower under realistic settings in which this bias is reduced. Such biases are most notably observed on datasets that are created by focused sampling instead of random sampling. Datasets with a higher proportion of implicit abuse are more affected than datasets with a lower proportion.
Recent work suggests that concreteness and imageability play an important role in the meanings of figurative expressions. We investigate this idea in several ways. First, we try to define more precisely the context within which a figurative expression may occur, by parsing a corpus annotated for metaphor. Next, we add both concreteness and imageability as “features” to the parsed metaphor corpus, by marking up words in this corpus using a psycholinguistic database of scores for concreteness and imageability. Finally, we carry out detailed statistical analyses of the augmented version of the original metaphor corpus, cross-matching the features of concreteness and imageability with others in the corpus such as parts of speech and dependency relations, in order to investigate in detail the use of such features in predicting whether a given expression is metaphorical or not.
This work proposes opinion frames as a representation of discourse-level associations which arise from related opinion topics. We illustrate how opinion frames help gather more information and also assist disambiguation. Finally we present the results of our experiments to detect these associations.
Discourse segmentation is the division of a text into minimal discourse segments, which form the leaves in the trees that are used to represent discourse structures. A definition of elementary discourse segments in German is provided by adapting widely used segmentation principles for English minimal units, while considering punctuation, morphology, sytax, and aspects of the logical document structure of a complex text type, namely scientific articles. The algorithm and implementation of a discourse segmenter based on these principles is presented, as well an evaluation of test runs.
The paper describes preliminary studies regarding the usage of Example-Based Querying for specialist corpora. We outline an infrastructure for its application within the linguistic domain. Example-Based Querying deals with retrieval situations where users would like to explore large collections of specialist texts semantically, but are unable to explicitly name the linguistic phenomenon they look for. As a way out, the proposed framework allows them to input prototypical everyday language examples or cases of doubt, which are automatically processed by CRF and linked to appropriate linguistic texts in the corpus.
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.
As many popular text genres such as blogs or news contain opinions by multiple sources and about multiple targets, finding the sources and targets of subjective expressions becomes an important sub-task for automatic opinion analysis systems. We argue that while automatic semantic role labeling systems (ASRL) have an important contribution to make, they cannot solve the problem for all cases. Based on the experience of manually annotating opinions, sources, and targets in various genres, we present linguistic phenomena that require knowledge beyond that of ASRL systems. In particular, we address issues relating to the attribution of opinions to sources; sources and targets that are realized as zero-forms; and inferred opinions. We also discuss in some depth that for arguing attitudes we need to be able to recover propositions and not only argued-about entities. A recurrent theme of the discussion is that close attention to specific discourse contexts is needed to identify sources and targets correctly.
In this paper, we present a suite of flexible UIMA-based components for information retrieval research which have been successfully used (and re-used) in several projects in different application domains. Implementing the whole system as UIMA components is beneficial for configuration management, component reuse, implementation costs, analysis and visualization.
We present a method and a software tool, the FrameNet Transformer, for deriving customized versions of the FrameNet database based on frame and frame element relations. The FrameNet Transformer allows users to iteratively coarsen the FrameNet sense inventory in two ways. First, the tool can merge entire frames that are related by user-specified relations. Second, it can merge word senses that belong to frames related by specified relations. Both methods can be interleaved. The Transformer automatically outputs format-compliant FrameNet versions, including modified corpus annotation files that can be used for automatic processing. The customized FrameNet versions can be used to determine which granularity is suitable for particular applications. In our evaluation of the tool, we show that our method increases accuracy of statistical semantic parsers by reducing the number of word-senses (frames) per lemma, and increasing the number of annotated sentences per lexical unit and frame. We further show in an experiment on the FATE corpus that by coarsening FrameNet we do not incur a significant loss of information that is relevant to the Recognizing Textual Entailment task.
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.
Accurate opinion mining requires the exact identification of the source and target of an opinion. To evaluate diverse tools, the research community relies on the existence of a gold standard corpus covering this need. Since such a corpus is currently not available for German, the Interest Group on German Sentiment Analysis decided to create such a resource and make it available to the research community in the context of a shared task. In this paper, we describe the selection of textual sources, development of annotation guidelines, and first evaluation results in the creation of a gold standard corpus for the German language.
In this paper we present work in developing a computerized grammar for the Latin language. It demonstrates the principles and challenges in developing a grammar for a natural language in a modern grammar formalism. The grammar presented here provides a useful resource for natural language processing applications in different fields. It can be easily adopted for language learning and use in language technology for Cultural Heritage like translation applications or to support post-correction of document digitization.
Implicitly abusive language – What does it actually look like and why are we not getting there?
(2021)
Abusive language detection is an emerging field in natural language processing which has received a large amount of attention recently. Still the success of automatic detection is limited. Particularly, the detection of implicitly abusive language, i.e. abusive language that is not conveyed by abusive words (e.g. dumbass or scum), is not working well. In this position paper, we explain why existing datasets make learning implicit abuse difficult and what needs to be changed in the design of such datasets. Arguing for a divide-and-conquer strategy, we present a list of subtypes of implicitly abusive language and formulate research tasks and questions for future research.
Semantic argument structures are often incomplete in that core arguments are not locally instantiated. However, many of these implicit arguments can be linked to referents in the wider context. In this paper we explore a number of linguistically motivated strategies for identifying and resolving such null instantiations (NIs). We show that a more sophisticated model for identifying definite NIs can lead to noticeable performance gains over the state-of-the- art for NI resolution.
Current Natural Language Processing (NLP) systems feature high-complexity processing pipelines that require the use of components at different levels of linguistic and application specific processing. These components often have to interface with external e.g. machine learning and information retrieval libraries as well as tools for human annotation and visualization. At the UKP Lab, we are working on the Darmstadt Knowledge Processing Software Repository (DKPro) (Gurevych et al., 2007a; Müller et al., 2008) to create a highly flexible, scalable and easy-to-use toolkit that allows rapid creation of complex NLP pipelines for semantic information processing on demand. The DKPro repository consists of several main parts created to serve the purposes of different NLP application areas
Integrated Linguistic Annotation Models and Their Application in the Domain of Antecedent Detection
(2011)
Seamless integration of various, often heterogeneous linguistic resources in terms of their output formats and a combined analysis of the respective annotation layers are crucial tasks for linguistic research. After a decade of concentration on the development of formats to structure single annotations for specific linguistic issues, in the last years a variety of specifications to store multiple annotations over the same primary data has been developed. The paper focuses on the integration of the knowledge resource logical document structure information into a text document to enhance the task of automatic anaphora resolution both for the task of candidate detection and antecedent selection. The paper investigates data structures necessary for knowledge integration and retrieval.
Natural language Processing tools are mostly developed for and optimized on newspaper texts, and often Show a substantial performance drop when applied to other types of texts such as Twitter feeds, Chat data or Internet forum posts. We explore a range of easy-to-implement methods of adapting existing part-of-speech taggers to improve their performance on Internet texts. Our results show that these methods can improve tagger performance substantially.
I’ve got a construction looks funny – representing and recovering non-standard constructions in UD
(2020)
The UD framework defines guidelines for a crosslingual syntactic analysis in the framework of dependency grammar, with the aim of providing a consistent treatment across languages that not only supports multilingual NLP applications but also facilitates typological studies. Until now, the UD framework has mostly focussed on bilexical grammatical relations. In the paper, we propose to add a constructional perspective and discuss several examples of spoken-language constructions that occur in multiple languages and challenge the current use of basic and enhanced UD relations. The examples include cases where the surface relations are deceptive, and syntactic amalgams that either involve unconnected subtrees or structures with multiply-headed dependents. We argue that a unified treatment of constructions across languages will increase the consistency of the UD annotations and thus the quality of the treebanks for linguistic analysis.
Linguistic query systems are special purpose IR applications. We present a novel state-of-the-art approach for the efficient exploitation of very large linguistic corpora, combining the advantages of relational database management systems (RDBMS) with the functional MapReduce programming model. Our implementation uses the German DEREKO reference corpus with multi-layer
linguistic annotations and several types of text-specific metadata, but the proposed strategy is language-independent and adaptable to large-scale multilingual corpora.
The use of digital resources and tools across humanities disciplines is steadily increasing, giving rise to new research paradigms and associated methods that are commonly subsumed under the term digital humanities. Digital humanities does not constitute a new discipline in itself, but rather a new approach to humanities research that cuts across different existing humanities disciplines. While digital humanities extends well beyond language-based research, textual resources and spoken language materials play a central role in most humanities disciplines.
We introduce a system that learns the participants of arbitrary given scripts. This system processes data from web experiments, in which each participant can be realized with different expressions. It computes participants by encoding semantic similarity and global structural information into an Integer Linear Program. An evaluation against a gold standard shows that we significantly outperform two informed baselines.
Linguistic Variation and Change in 250 Years of English Scientific Writing: A Data-Driven Approach
(2020)
We trace the evolution of Scientific English through the Late Modern period to modern time on the basis of a comprehensive corpus composed of the Transactions and Proceedings of the Royal Society of London, the first and longest-running English scientific journal established in 1665. Specifically, we explore the linguistic imprints of specialization and diversification in the science domain which accumulate in the formation of “scientific language” and field-specific sublanguages/registers (chemistry, biology etc.). We pursue an exploratory, data-driven approach using state-of-the-art computational language models and combine them with selected information-theoretic measures (entropy, relative entropy) for comparing models along relevant dimensions of variation (time, register). Focusing on selected linguistic variables (lexis, grammar), we show how we deploy computational language models for capturing linguistic variation and change and discuss benefits and limitations.
This paper presents first steps towards metaphor detection in German poetry, in particular in expressionist poems. We create a dataset with adjective-noun pairs extracted from expressionist poems, manually annotated for metaphoricity. We discuss the annotation process and present models and experiments for metaphor detection where we investigate the impact of context and the domain dependence of the models.
We present the annotation of information structure in the MULI project. To learn more about the information structuring means in prosody, syntax and discourse, theory- independent features were defined for each level. We describe the features and illustrate them on an example sentence. To investigate the interplay of features, the representation has to allow for inspecting all three layers at the same time. This is realised by a stand-off XML mark-up with the word as the basic unit. The theory-neutral XML stand-off annotation allows integrating this resource with other linguistic resources such as the Tiger Treebank for German or the Penn treebank for English.
This article introduces the topic of ‘‘Multilingual language resources and interoperability’’. We start with a taxonomy and parameters for classifying language resources. Later we provide examples and issues of interoperatability, and resource architectures to solve such issues. Finally we discuss aspects of linguistic formalisms and interoperability.
We present an approach for modeling German negation in open-domain fine grained sentiment analysis. Unlike most previous work in sentiment analysis, we assume that negation can be conveyed by many lexical units (and not only common negation words) and that different negation words have different scopes. Our approach is examined on a new dataset comprising sentences with mentions of polar expressions and various negation words. We identify different types of negation words that have the same scopes. We show that already negation modeling based on these types largely outperforms traditional negation models which assume the same scope for all negation words and which employ a window-based scope detection rather than a scope detection based on syntactic information.
The demo presents a minimalist, off-the-shelf AND tool which provides a fundamental AND operation, the comparison of two publications with ambiguous authors, as an easily accessible HTTP interface. The tool implements this operation using standard AND functionality, but puts particular emphasis on advanced methods from natural language processing (NLP) for comparing publication title semantics.
We present an approach for opinion role induction for verbal predicates. Our model rests on the assumption that opinion verbs can be divided into three different types where each type is associated with a characteristic mapping between semantic roles and opinion holders and targets. In several experiments, we demonstrate the relevance of those three categories for the task. We show that verbs can easily be categorized with semi-supervised graphbased clustering and some appropriate similarity metric. The seeds are obtained through linguistic diagnostics. We evaluate our approach against a new manually-compiled opinion role lexicon and perform in-context classification.