Refine
Year of publication
Document Type
- Conference Proceeding (688) (remove)
Keywords
- Korpus <Linguistik> (237)
- Deutsch (167)
- Computerlinguistik (117)
- Annotation (65)
- Automatische Sprachanalyse (53)
- Gesprochene Sprache (53)
- Natürliche Sprache (41)
- Forschungsdaten (38)
- Information Extraction (30)
- Metadaten (30)
Publicationstate
- Veröffentlichungsversion (442)
- Zweitveröffentlichung (81)
- Postprint (38)
- Preprint (1)
Reviewstate
- Peer-Review (328)
- (Verlags)-Lektorat (137)
- Peer-review (9)
- Review-Status-unbekannt (7)
- Peer review (1)
- Verlags-Lektorat (1)
Publisher
- European Language Resources Association (ELRA) (50)
- Association for Computational Linguistics (43)
- European Language Resources Association (35)
- Institut für Deutsche Sprache (17)
- Zenodo (15)
- Lexical Computing CZ s.r.o. (12)
- Linköping University Electronic Press (12)
- CLARIN (11)
- International Speech Communication Association (9)
- Leibniz-Institut für Deutsche Sprache (9)
The Component MetaData Infrastructure (CMDI) is a framework for the creation and usage of metadata formats to describe all kinds of resources in the CLARIN world. To better connect to the library world, and to allow librarians to enter metadata for linguistic resources into their catalogues, a crosswalk from CMDI-based formats to bibliographic standards is required. The general and rather fluid nature of CMDI, however, makes it hard to map arbitrary CMDI schemas to metadata standards such as Dublin Core (DC) or MARC 21, which have a mature, well-defined and fixed set of field descriptors. In this paper, we address the issue and propose crosswalks between CMDI-based profiles originating from the NaLiDa project and DC and MARC 21, respectively.
The ISOcat registry reloaded
(2012)
The linguistics community is building a metadata-based infrastructure for the description of its research data and tools. At its core is the ISOcat registry, a collaborative platform to hold a (to be standardized) set of data categories (i.e., field descriptors). Descriptors have definitions in natural language and little explicit interrelations. With the registry growing to many hundred entries, authored by many, it is becoming increasingly apparent that the rather informal definitions and their glossary-like design make it hard for users to grasp, exploit and manage the registry’s content. In this paper, we take a large subset of the ISOcat term set and reconstruct from it a tree structure following the footsteps of schema.org. Our ontological re-engineering yields a representation that gives users a hierarchical view of linguistic, metadata-related terminology. The new representation adds to the precision of all definitions by making explicit information which is only implicitly given in the ISOcat registry. It also helps uncovering and addressing potential inconsistencies in term definitions as well as gaps and redundancies in the overall ISOcat term set. The new representation can serve as a complement to the existing ISOcat model, providing additional support for authors and users in browsing, (re-)using, maintaining, and further extending the community’s terminological metadata repertoire.
This article presents preliminary results indicating that speakers have a different pitch range when they speak a foreign language compared to the pitch variation that occurs when they speak their native language. To this end, a learner corpus with French and German speakers was analyzed. Results suggest that speakers indeed produce a smaller pitch range in the respective L2. This is true for both groups of native speakers. A possible explanation for this finding is that speakers are less confident in their productions, therefore, they concentrate more on segments and words and subsequently refrain from realizing pitch range more native-like. For language teaching, the results suggest that learners should be trained extensively on the more pronounced use of pitch in the foreign language.
This study examines the pitch profiles of French learners of German and German learners of French, both in their native language (L1), and in their respective foreign language (L2). Results of the analysis of 84 speakers suggest that for short read sentences, French and German speakers do not show pitch range differences in their native production. Furthermore, analyses of mean f0 and pitch range indicate that range is not necessarily reduced in L2 productions. These results are different from results reported in prior research. Possible reasons for these differences are discussed.
Smooth turn-taking in conversation depends in part on speakers being able to communicate their intention to hold or cede the floor. Both prosodic and gestural cues have been shown to be used in this context. We investigate the interplay of pitch movements and hand gestures at locations at which speaker change becomes relevant, comparing their use in German and Swedish. We find that there are some shared functions of prosody and gesture with regard to turn-taking in the two languages, but that these shared functions appear to be mediated by the different phonological demands on pitch in the two languages.
Linguistic corpora have been annotated by means of SGML-based markup languages for almost 20 years. We can, very roughly, differentiate between three distinct evolutionary stages of markup technologies. (1)Originally, single SGML tree-based document instances were deemed sufficient for the representation of linguistic structures. (2) Linguists began to realize that alternatives and extensions to the traditional model are needed. Formalisms such as, for example, NITE were proposed: the NITE Object Model (NOM) consists of multi-rooted trees. (3) We are now on the threshold of the third evolutionary stage: even NITE's very flexible approach is not suited for all linguistic purposes. As some structures, such as these, cannot be modeled by multi-rooted trees, an even more flexible approach is needed in order to provide a generic annotation format that is able to represent genuinely arbitrary linguistic data structures.
Accentuation, Uncertainty and Exhaustivity - Towards a Model of Pragmatic Focus Interpretation
(2010)
This paper presents a model of pragmatic focus interpretation that is assumed to be part of a complete language comprehension model and that is inspired by Levelt's language processing model. The model is derived from our empirical data on the role of accentuation, prosodic indicators of uncertainty and context for pragmatic focus interpretation. In its present state, the model is restricted to these data, but nevertheless generates predictions.
We start by trying to answer a question that has already been asked by de Schryver et al. (2006): Do dictionary users (frequently) look up words that are frequent in a corpus. Contrary to their results, our results that are based on the analysis of log files from two different online dictionaries indicate that users indeed look up frequent words frequently. When combining frequency information from the Mannheim German Reference Corpus and information about the number of visits in the Digital Dictionary of the German Language as well as the German language edition of Wiktionary, a clear connection between corpus and look-up frequencies can be observed. In a follow-up study, we show that another important factor for the look-up frequency of a word is its temporal social relevance. To make this effect visible, we propose a de-trending method where we control both frequency effects and overall look-up trends.
Reading corpora are text collections that are enriched with processing data. From a corpus linguist’s perspective, they can be seen as an extension of classical linguistic corpora with human language processing behavior. From a psycholinguist’s perspective, reading corpora allow to test psycholinguistic hypotheses on subsets of language and language processing as it is ‘in the wild’ – in contrast to strictly controlled language material in isolated sentences, as used in most psycholinguistic experiments. In this paper, we will investigate a relevance-based account of language processing which states that linguistic structures, that are embedded deeper syntactically, are read faster because readers allocate less attention to these structures.
On the Lossless Transformation of Single-File, Multi-Layer Annotations into Multi-Rooted Trees
(2007)
The Generalised Architecture for Sustainability (GENAU) provides a framework for the transformation of single-file, multi-layer annotations into multi-rooted trees. By employing constraints expressed in XCONCUR-CL, this procedure can be performed lossless, i.e., without losing information, especially with regard to the nesting of elements that belong to multiple annotation layers. This article describes how different types of linguistic corpora can be transformed using specialised tools, and how constraint rules can be applied to the resulting multi-rooted trees to add an additional level of validation.
We describe a general two-stage procedure for re-using a custom corpus for spoken language system development involving a transformation from character-based markup to XML, and DSSSL stylesheet-driven XML markup enhancement with multiple lexical tag trees. The procedure was used to generate a fully tagged corpus; alternatively with greater economy of computing resources, it can be employed as a parametrised ‘tagging on demand’ filter. The implementation will shortly be released as a public resource together with the corpus (German spoken dialogue, about 500k word form tokens) and lexicon (about 75k word form types).
The Leibniz-Institute for the German Language (IDS) was established in Mannheim in 1964. Since then, it has been at the forefront of innovation in German linguistics as a hub for digital language data. This chapter presents various lessons learnt from over five decades of work by the IDS, ranging from the importance of sustainability, through its strong technical base and FAIR principles, to the IDS’ role in national and international cooperation projects and its expertise on legal and ethical issues related to language resources and language technology.
Overlap in markup occurs where some markup structures do not nest, such as where the structural division of the text into lists, sections, etc., differs from the syntactic division of the text into sentences and phrases. The Multiple Annotation solution to this problem (redundant encoding in multiple forms) has many advantages: it is based on XML, the modeling of alternative annotations is possible, each level can be viewed separately, and new levels can be added at any time. But it has the significant disadvantage of independence of the separate files. These multiply annotated files can be regarded as an interrelated unit, with the text serving as the implicit link. Two representations of the information contained in the multiple files (one in Prolog and one in XML) can be programmatically derived and used together for editing, for inference, or for unification of the multiply annotated documents.
This paper describes work directed towards the development of a syllable prominence-based prosody generation functionality for a German unit selection speech synthesis system. A general concept for syllable prominence-based prosody generation in unit selection synthesis is proposed. As a first step towards its implementation, an automated syllable prominence annotation procedure based on acoustic analyses has been performed on the BOSS speech corpus. The prominence labeling has been evaluated against an existing annotation of lexical stress levels and manual prominence labeling on a subset of the corpus. We discuss methods and results and give an outlook on further implementation steps.
The present study introduces articulography, the measurement of the position of tongue and lips during speech, as a promising method to the study of dialect variation. By using generalized additive modeling to analyze articulatory trajectories, we are able to reliably detect aggregate group differences, while simultaneously taking into account the individual variation across dozens of speakers. Our results on the basis of Dutch dialect data show clear differences between the southern and the northern dialect with respect to tongue position, with a more frontal tongue position in the dialect from Ubbergen (in the southern half of the Netherlands) than in the dialect of Ter Apel (in the northern half of the Netherlands). Thus articulography appears to be a suitable tool to investigate structural differences in pronunciation at the dialect level.
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.
We examine different features and classifiers for the categorization of opinion words into actor and speaker view. To our knowledge, this is the first comprehensive work to address sentiment views on the word level taking into consideration opinion verbs, nouns and adjectives. We consider many high-level features requiring only few labeled training data. A detailed feature analysis produces linguistic insights into the nature of sentiment views. We also examine how far global constraints between different opinion words help to increase classification performance. Finally, we show that our (prior) word-level annotation correlates with contextual sentiment views.
Opinion Holder and Target Extraction for Verb-based Opinion Predicates – The Problem is Not Solved
(2015)
We offer a critical review of the current state of opinion role extraction involving opinion verbs. We argue that neither the currently available lexical resources nor the manually annotated text corpora are sufficient to appropriately study this task. We introduce a new corpus focusing on opinion roles of opinion verbs from the Subjectivity Lexicon and show potential benefits of this corpus. We also demonstrate that state-of-the-art classifiers perform rather poorly on this new dataset compared to the standard dataset for the task showing that there still remains significant research to be done.
We address the detection of abusive words. The task is to identify such words among a set of negative polar expressions. We propose novel features employing information from both corpora and lexical resources. These features are calibrated on a small manually annotated base lexicon which we use to produce a large lexicon. We show that the word-level information we learn cannot be equally derived from a large dataset of annotated microposts. We demonstrate the effectiveness of our (domain-independent) lexicon in the crossdomain detection of abusive microposts.
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.
We examine predicative adjectives as an unsupervised criterion to extract subjective adjectives. We do not only compare this criterion with a weakly supervised extraction method but also with gradable adjectives, i.e. another highly subjective subset of adjectives that can be extracted in an unsupervised fashion. In order to prove the robustness of this extraction method, we will evaluate the extraction with the help of two different state-of-the-art sentiment lexicons (as a gold standard).
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.
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.
We propose to use abusive emojis, such as the “middle finger” or “face vomiting”, as a proxy for learning a lexicon of abusive words. Since it represents extralinguistic information, a single emoji can co-occur with different forms of explicitly abusive utterances. We show that our approach generates a lexicon that offers the same performance in cross-domain classification of abusive microposts as the most advanced lexicon induction method. Such an approach, in contrast, is dependent on manually annotated seed words and expensive lexical resources for bootstrapping (e.g. WordNet). We demonstrate that the same emojis can also be effectively used in languages other than English. Finally, we also show that emojis can be exploited for classifying mentions of ambiguous words, such as “fuck” and “bitch”, into generally abusive and just profane usages.
We present a gold standard for semantic relation extraction in the food domain for German. The relation types that we address are motivated by scenarios for which IT applications present a commercial potential, such as virtual customer advice in which a virtual agent assists a customer in a supermarket in finding those products that satisfy their needs best. Moreover, we focus on those relation types that can be extracted from natural language text corpora, ideally content from the internet, such as web forums, that are easy to retrieve. A typical relation type that meets these requirements are pairs of food items that are usually consumed together. Such a relation type could be used by a virtual agent to suggest additional products available in a shop that would potentially complement the items a customer has already in their shopping cart. Our gold standard comprises structural data, i.e. relation tables, which encode relation instances. These tables are vital in order to evaluate natural language processing systems that extract those relations.
Knowledge Acquisition with Natural Language Processing in the Food Domain: Potential and Challenges
(2012)
In this paper, we present an outlook on the effectiveness of natural language processing (NLP) in extracting knowledge for the food domain. We identify potential scenarios that we think are particularly suitable for NLP techniques. As a source for extracting knowledge we will highlight the benefits of textual content from social media. Typical methods that we think would be suitable will be discussed. We will also address potential problems and limits that the application of NLP methods may yield.
In this paper, we examine methods to automatically extract domain-specific knowledge from the food domain from unlabeled natural language text. We employ different extraction methods ranging from surface patterns to co-occurrence measures applied on different parts of a document. We show that the effectiveness of a particular method depends very much on the relation type considered and that there is no single method that works equally well for every relation type. We also examine a combination of extraction methods and also consider relationships between different relation types. The extraction methods are applied both on a domain-specific corpus and the domain-independent factual knowledge base Wikipedia. Moreover, we examine an open-domain lexical ontology for suitability.
Automatic Food Categorization from Large Unlabeled Corpora and Its Impact on Relation Extraction
(2014)
We present a weakly-supervised induction method to assign semantic information to food items. We consider two tasks of categorizations being food-type classification and the distinction of whether a food item is composite or not. The categorizations are induced by a graph-based algorithm applied on a large unlabeled domain-specific corpus. We show that the usage of a domain-specific corpus is vital. We do not only outperform a manually designed open-domain ontology but also prove the usefulness of these categorizations in relation extraction, outperforming state-of-the-art features that include syntactic information and Brown clustering.
We examine the combination of pattern-based and distributional similarity for the induction of semantic categories. Pattern-based methods are precise and sparse while distributional methods have a higher recall. Given these particular properties we use the prediction of distributional methods as a back-off to pattern-based similarity. Since our pattern-based approach is embedded into a semi-supervised graph clustering algorithm, we also examine how distributional information is best added to that classifier. Our experiments are carried out on 5 different food categorization tasks.
In this paper, we examine methods to extract different domain-specific relations from the food domain. We employ different extraction methods ranging from surface patterns to co-occurrence measures applied on different parts of a document. We show that the effectiveness of a particular method depends very much on the relation type considered and that there is no single method that works equally well for every relation type. As we need to process a large amount of unlabeled data our methods only require a low level of linguistic processing. This has also the advantage that these methods can provide responses in real time.
One problem of data-driven answer extraction in open-domain factoid question answering is that the class distribution of labeled training data is fairly imbalanced. In an ordinary training set, there are far more incorrect answers than correct answers. The class-imbalance is, thus, inherent to the classification task. It has a deteriorating effect on the performance of classifiers trained by standard machine learning algorithms. They usually have a heavy bias towards the majority class, i.e. the class which occurs most often in the training set. In this paper, we propose a method to tackle class imbalance by applying some form of cost-sensitive learning which is preferable to sampling. We present a simple but effective way of estimating the misclassification costs on the basis of class distribution. This approach offers three benefits. Firstly, it maintains the distribution of the classes of the labeled training data. Secondly, this form of meta-learning can be applied to a wide range of common learning algorithms. Thirdly, this approach can be easily implemented with the help of state-of-the-art machine learning software.
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.
We examine the task of separating types from brands in the food domain. Framing the problem as a ranking task, we convert simple textual features extracted from a domain-specific corpus into a ranker without the need of labeled training data. Such method should rank brands (e.g. sprite) higher than types (e.g. lemonade). Apart from that, we also exploit knowledge induced by semi-supervised graph-based clustering for two different purposes. On the one hand, we produce an auxiliary categorization of food items according to the Food Guide Pyramid, and assume that a food item is a type when it belongs to a category unlikely to contain brands. On the other hand, we directly model the task of brand detection using seeds provided by the output of the textual ranking features. We also harness Wikipedia articles as an additional knowledge source.
We explore the feasibility of contextual healthiness classification of food items. We present a detailed analysis of the linguistic phenomena that need to be taken into consideration for this task based on a specially annotated corpus extracted from web forum entries. For automatic classification, we compare a supervised classifier and rule-based classification. Beyond linguistically motivated features that include sentiment information we also consider the prior healthiness of food items.
In this paper, we explore different linguistic structures encoded as convolution kernels for the detection of subjective expressions. The advantage of convolution kernels is that complex structures can be directly provided to a classifier without deriving explicit features. The feature design for the detection of subjective expressions is fairly difficult and there currently exists no commonly accepted feature set. We consider various structures, such as constituency parse structures, dependency parse structures, and predicate-argument structures. In order to generalize from lexical information, we additionally augment these structures with clustering information and the task-specific knowledge of subjective words. The convolution kernels will be compared with a standard vector kernel.
In recent years, text classification in sentiment analysis has mostly focused on two types of classification, the distinction between objective and subjective text, i.e. subjectivity detection, and the distinction between positive and negative subjective text, i.e. polarity classification. So far, there has been little work examining the distinction between definite polar subjectivity and indefinite polar subjectivity. While the former are utterances which can be categorized as either positive or negative, the latter cannot be categorized as either of these two categories. This paper presents a small set of domain independent features to detect indefinite polar sentences. The features reflect the linguistic structure underlying these types of utterances. We give evidence for the effectiveness of these features by incorporating them into an unsupervised rule-based classifier for sentence-level analysis and compare its performance with supervised machine learning classifiers, i.e. Support Vector Machines (SVMs) and Nearest Neighbor Classifier (kNN). The data used for the experiments are web-reviews collected from three different domains.
In order to automatically extract opinion holders, we propose to harness the contexts of prototypical opinion holders, i.e. common nouns, such as experts or analysts, that describe particular groups of people whose profession or occupation is to form and express opinions towards specific items. We assess their effectiveness in supervised learning where these contexts are regarded as labelled training data and in rule-based classification which uses predicates that frequently co-occur with mentions of the prototypical opinion holders. Finally, we also examine in how far knowledge gained from these contexts can compensate the lack of large amounts of labeled training data in supervised learning by considering various amounts of actually labeled training sets.
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.
In this paper, we investigate the role of predicates in opinion holder extraction. We will examine the shape of these predicates, investigate what relationship they bear towards opinion holders, determine what resources are potentially useful for acquiring them, and point out limitations of an opinion holder extraction system based on these predicates. For this study, we will carry out an evaluation on a corpus annotated with opinion holders. Our insights are, in particular, important for situations in which no labelled training data are available and only rule-based methods can be applied.
Bootstrapping Supervised Machine-learning Polarity Classifiers with Rule-based Classification
(2010)
In this paper, we explore the effectiveness of bootstrapping supervised machine-learning polarity classifiers using the output of domain-independent rule-based classifiers. The benefit of this method is that no labeled training data are required. Still, this method allows to capture in-domain knowledge by training the supervised classifier on in-domain features, such as bag of words.
We investigate how important the quality of the rule-based classifier is and what features are useful for the supervised classifier. The former addresses the issue in how far relevant constructions for polarity classification, such as word sense disambiguation, negation modeling, or intensification, are important for this self-training approach. We not only compare how this method relates to conventional semi-supervised learning but also examine how it performs under more difficult settings in which classes are not balanced and mixed reviews are included in the dataset.
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.
Opinion holder extraction is one of the important subtasks in sentiment analysis. The effective detection of an opinion holder depends on the consideration of various cues on various levels of representation, though they are hard to formulate explicitly as features. In this work, we propose to use convolution kernels for that task which identify meaningful fragments of sequences or trees by themselves. We not only investigate how different levels of information can be effectively combined in different kernels but also examine how the scope of these kernels should be chosen. In general relation extraction, the two candidate entities thought to be involved in a relation are commonly chosen to be the boundaries of sequences and trees. The definition of boundaries in opinion holder extraction, however, is less straightforward since there might be several expressions beside the candidate opinion holder to be eligible for being a boundary.
In this paper, we compare three different generalization methods for in-domain and cross-domain opinion holder extraction being simple unsupervised word clustering, an induction method inspired by distant supervision and the usage of lexical resources. The generalization methods are incorporated into diverse classifiers. We show that generalization causes significant improvements and that the impact of improvement depends on the type of classifier and on how much training and test data differ from each other. We also address the less common case of opinion holders being realized in patient position and suggest approaches including a novel (linguistically-informed) extraction method how to detect those opinion holders without labeled training data as standard datasets contain too few instances of this type.