Refine
Year of publication
Document Type
- Conference Proceeding (236) (remove)
Language
- English (236) (remove)
Has Fulltext
- yes (236)
Keywords
- Korpus <Linguistik> (100)
- Computerlinguistik (58)
- Deutsch (43)
- Annotation (27)
- Forschungsdaten (25)
- Datenmanagement (24)
- Automatische Sprachanalyse (23)
- Metadaten (21)
- Gesprochene Sprache (19)
- Natürliche Sprache (19)
Publicationstate
- Veröffentlichungsversion (236) (remove)
Reviewstate
- Peer-Review (236) (remove)
Publisher
- Association for Computational Linguistics (31)
- European Language Resources Association (17)
- Institut für Deutsche Sprache (17)
- European Language Resources Association (ELRA) (14)
- Lexical Computing CZ s.r.o. (12)
- Linköping University Electronic Press (12)
- CLARIN (10)
- Leibniz-Institut für Deutsche Sprache (9)
- German Society for Computational Linguistics & Language Technology und Friedrich-Alexander-Universität Erlangen-Nürnberg (6)
- The Association for Computational Linguistics (5)
The paper reports the results of the curation project ChatCorpus2CLARIN. The goal of the project was to develop a workflow and resources for the integration of an existing chat corpus into the CLARIN-D research infrastructure for language resources and tools in the Humanities and the Social Sciences (http://clarin-d.de). The paper presents an overview of the resources and practices developed in the project, describes the added value of the resource after its integration and discusses, as an outlook, to what extent these practices can be considered best practices which may be useful for the annotation and representation of other CMC and social media corpora.
Song lyrics can be considered as a text genre that has features of both written and spoken discourse, and potentially provides extensive linguistic and cultural information to scientists from various disciplines. However, pop songs play a rather subordinate role in empirical language research so far - most likely due to the absence of scientifically valid and sustainable resources. The present paper introduces a multiply annotated corpus of German lyrics as a publicly available basis for multidisciplinary research. The resource contains three types of data for the investigation and evaluation of quite distinct phenomena: TEI-compliant song lyrics as primary data, linguistically and literary motivated annotations, and extralinguistic metadata. It promotes empirically/statistically grounded analyses of genre-specific features, systemic-structural correlations and tendencies in the texts of contemporary pop music. The corpus has been stratified into thematic and author-specific archives; the paper presents some basic descriptive statistics, as well as the public online frontend with its built-in evaluation forms and live visualisations.
In this paper, we will present a first attempt to classify commonly confused words in German by consulting their communicative functions in corpora. Although the use of so-called paronyms causes frequent uncertainties due to similarities in spelling, sound and semantics, up until now the phenomenon has attracted little attention either from the perspective of corpus linguistics or from cognitive linguistics. Existing investigations rely on structuralist models, which do not account for empirical evidence. Still, they have developed an elaborate model based on formal criteria, primarily on word formation (cf. Lăzărescu 1999). Looking from a corpus perspective, such classifications are incompatible with language in use and cognitive elements of misuse.
This article sketches first lexicological insights into a classification model as derived from semantic analyses of written communication. Firstly, a brief description of the project will be provided. Secondly, corpus-assisted paronym detection will be focused. Thirdly, in the main section the paper concerns the description of the datasets for paronym classification and the classification procedures. As a work in progress, new insights will continually be extended once spoken and CMC data are added to the investigations.
This paper presents the prototype of a lexicographic resource for spoken German in interaction, which was conceived within the framework of the LeGeDe-project (LeGeDe=Lexik des gesprochenen Deutsch). First of all, it summarizes the theoretical and methodological approaches that were used for the initial planning of the resource. The headword candidates were selected by analyzing corpus-based data. Therefore, the data of two corpora (written and spoken German) were compared with quantitative methods. The information that was gathered on the selected headword candidates can be assigned to two different sections: meanings and functions in interaction.
Additionally, two studies on the expectations of future users towards the resource were carried out. The results of these two studies were also taken into account in the development of the prototype. Focusing on the presentation of the resource’s content, the paper shows both the different lexicographical information in selected dictionary entries, and the information offered by the provided hyperlinks and external texts. As a conclusion, it summarizes the most important innovative aspects that were specifically developed for the implementation of such a resource.
Ph@ttSessionz and Deutsch heute are two large German speech databases. They were created for different purposes: Ph@ttSessionz to test Internet-based recordings and to adapt speech recognizers to the voices of adolescent speakers, Deutsch heute to document regional variation of German. The databases differ in their recording technique, the selection of recording locations and speakers, elicitation mode, and data processing.
In this paper, we outline how the recordings were performed, how the data was processed and annotated, and how the two databases were imported into a single relational database system. We present acoustical measurements on the digit items of both databases. Our results confirm that the elicitation technique affects the speech produced, that f0 is quite comparable despite different recording procedures, and that large speech technology databases with suitable metadata may well be used for the analysis of regional variation of speech.
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 present an implemented XML data model and a new, simplified query language for multi-level annotated corpora. The new query language involves automatic conversion of queries into the underlying, more complicated MMAXQL query language. It supports queries for sequential and hierarchical, but also associative (e.g. coreferential) relations. The simplified query language has been designed with non-expert users in mind.
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.
We present a testsuite for POS tagging German web data. Our testsuite provides the original raw text as well as the gold tokenisations and is annotated for parts-of-speech. The testsuite includes a new dataset for German tweets, with a current size of 3,940 tokens. To increase the size of the data, we harmonised the annotations in already existing web corpora, based on the Stuttgart-Tübingen Tag Set. The current version of the corpus has an overall size of 48,344 tokens of web data, around half of it from Twitter. We also present experiments, showing how different experimental setups (training set size, additional out-of-domain training data, self-training) influence the accuracy of the taggers. All resources and models will be made publicly available to the research community.
We apply a decision tree based approach to pronoun resolution in spoken dialogue. Our system deals with pronouns with NP- and non-NP-antecedents. We present a set of features designed for pronoun resolution in spoken dialogue and determine the most promising features. We evaluate the system on twenty Switchboard dialogues and show that it compares well to Byron’s (2002) manually tuned system.
Creating and maintaining metadata for various kinds of resources requires appropriate tools to assist the user. The paper presents the metadata editor ProFormA for the creation and editing of CMDI (Component Metadata Infrastructure) metadata in web forms. This editor supports a number of CMDI profiles currently being provided for different types of resources. Since the editor is based on XForms and server-side processing, users can create and modify CMDI files in their standard browser without the need for further processing. Large parts of ProFormA are implemented as web services in order to reuse them in other contexts and programs.
The understanding of story variation, whether motivated by cultural currents or other factors, is important for applications of formal models of narrative such as story generation or story retrieval. We present the first stage of an experiment to elicit natural narrative variation data suitable for evaluation with respect to story similarity, to qualitative and quantitative analysis of story variation, and also for data processing. We also present few preliminary results from the first stage of the experiment, using Red Riding Hood and Romeo and Juliet as base texts.
XML has been designed for creating structured documents, but the information that is encoded in these structures are, by definition, out of scope for XML. Additional sources, normally not easily interpretable by computers, such as documentation are needed to determine the intention of specific tags in a tag-set. The Component Metadata Infrastructure (CMDI) takes a rather pragmatic approach to foster interoperability between XML instances in the domain of metadata descriptions for language resources. This paper gives an overview of this approach.
This paper presents the system architecture as well as the underlying workflow of the Extensible Repository System of Digital Objects (ERDO) which has been developed for the sustainable archiving of language resources within the Tübingen CLARIN-D project. In contrast to other approaches focusing on archiving experts, the described workflow can be used by researchers without required knowledge in the field of long-term storage for transferring data from their local file systems into a persistent repository.
This paper describes the lexical database tool LOLA (Linguistic-Oriented Lexical database Approach) which has been developed for the construction and maintenance of lexicons for the machine translation system LMT. First, the requirements such a tool should meet are discussed, then LMT and the lexical information it requires, and some issues concerning vocabulary acquisition are presented. Afterwards the architecture and the components of the LOLA system are described and it is shown how we tried to meet the requirements worked out earlier. Although LOLA originally has been designed and implemented for the German-English LMT prototype, it aimed from the beginning at a representation of lexical data that can be reused for other LMT or MT prototypes or even other NLP applications. A special point of discussion will therefore be the adaptability of the tool and its components as well as the reusability of the lexical data stored in the database for the lexicon development for LMT or for other applications.
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 paper presents a survey on hate speech detection. Given the steadily growing body of social media content, the amount of online hate speech is also increasing. Due to the massive scale of the web, methods that automatically detect hate speech are required. Our survey describes key areas that have been explored to automatically recognize these types of utterances using natural language processing. We also discuss limits of those approaches.
This paper presents a survey on the role of negation in sentiment analysis. Negation is a very common linguistic construction that affects polarity and, therefore, needs to be taken into consideration in sentiment analysis.
We will present various computational approaches modeling negation in sentiment analysis. We will, in particular, focus on aspects such as level of representation used for sentiment analysis, negation word detection and scope of negation. We will also discuss limits and challenges of negation modeling on that task.
The Lehnwortportal Deutsch (2012 seqq.) serves as an integrated online information system on German lexical borrowings into other languages, synthesizing an increasing number of lexicographical dictionaries and providing basic cross-resource search options. The paper discusses the far-reaching revision of the system’s conceptual, lexicographical and technological underpinnings currently under way, focussing on their relevance for multilingual loanword lexicography.
In this paper we present an experimental semantic search function, based on word embeddings, for an integrated online information system on German lexical borrowings into other languages, the Lehnwortportal Deutsch (LWPD). The LWPD synthesizes an increasing number of lexicographical resources and provides basic cross-resource search options. Onomasiological access to the lexical units of the portal is a highly desirable feature for many research questions, such as the likelihood of borrowing lexical units with a given meaning (Haspelmath & Tadmor, 2009; Zeller, 2015). The search technology is based on multilingual pre-trained word embeddings, and individual word senses in the portal are associated with word vectors. Users may select one or more among a very large number of search terms, and the database returns lexical items with word sense vectors similar to these terms. We give a preliminary assessment of the feasibility, usability and efficacy of our approach, in particular in comparison to search options based on semantic domains or fields.
The Manatee corpus management system on which the Sketch Engine is built is efficient, but unable to harness the power of today’s multiprocessor machines. We describe a new, compatible implementation of Manatee which we develop in the Go language and report on the performance gains that we obtained.
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.
This paper addresses long-term archival for large corpora. Three aspects specific to language resources are focused, namely (1) the removal of resources for legal reasons, (2) versioning of (unchanged) objects in constantly growing resources, especially where objects can be part of multiple releases but also part of different collections, and (3) the conversion of data to new formats for digital preservation. It is motivated why language resources may have to be changed, and why formats may need to be converted. As a solution, the use of an intermediate proxy object called a signpost is suggested. The approach will be exemplified with respect to the corpora of the Leibniz Institute for the German Language in Mannheim, namely the German Reference Corpus (DeReKo) and the Archive for Spoken German (AGD).
Besides English, Afrikaans is considered “the [Germanic] language which deviates grammatically the farthest from the others” (Harbert 2007: 17). But how exactly do we measure “grammatical deviation”, and how deviant is Afrikaans really if we compare it not just to other standard languages but also to non-standard varieties? The present contribution aims to address those questions combining functional-typological and dialectometric perspectives. We first select data for 28 Germanic varieties showing vastly different speaker numbers, grades of standardisation and amounts of language contact. Based on 48 (micro)typological variables from syntax, morphology and phonology, we perform cluster analysis and multidimensional scaling and present ways of visualizing and interpreting the results. Inter alia, the analyses show a major divide between Continental West Germanic and North Germanic (as might be expected) and they also identify a number of outliers, including English and pidgin and creole languages such as Russenorsk or Rabaul Creole German. Afrikaans appears to cluster with the other West Germanic languages rather than the outliers. Within West Germanic, however, it does indeed emerge as rather deviant and, according to our metric, it is, for example, typologically closer to other high-contact varieties such as Yiddish than it is to Dutch.
We describe a simple and efficient Java object model and application programming interface (API) for (possibly multi-modal) annotated natural language corpora. Corpora are represented as elements like Sentences, Turns, Utterances, Words, Gestures and Markables. The API allows linguists to access corpora in terms of these discourse-level elements, i.e. at a conceptual level they are familiar with, with the flexibility offered by a general purpose programming language. It is also a contribution to corpus standardization efforts because it is based on a straightforward and easily extensible data model which can serve as a target for conversion of different corpus formats.
We investigate whether prototypicality or prominence of semantic roles can account for role-related effects in sentence interpretation. We present two acceptability-rating experiments testing three different constructions: active, personal passive and DO-clefts involving the same type of transitive verbs that differ with respect to the agentive role features they select. Our results reveal that there is no cross-constructional advantage for prototypical roles (e.g., agents), hence disconfirming a central tenet of role prototypicality. Rather, acceptability clines depend on the construction under investigation, thereby highlighting different role features. This finding is in line with one core assumption of the prominence account stating that role features are flexibly highlighted depending on the discourse function of the respective construction.
This paper aims to address these problems by dealing with theoretical and methodological questions concerning the national effects of the Bologna Process and the role national factors play in determining the impact of these effects. Altogether the purpose of the paper is to serve as a starting point for future research – both as a guide for systematic and comparative empirical work on higher education, but also for further theoretical and methodological reasoning concerning research on (higher) education policy. As higher education research so far particularly lacks an approach allowing for a competitive and systematic falsification of theoretical arguments by clearly indicating testable and specific hypothesis as well as variables behind the research design (Goedegebuure/Vught 1996) we propose to fall back on neighbouring disciplines, namely social science to improve and enhance the analysis (Slaughter 2001: 398; Altbach 2002: 154; Teichler 1996a: 433, 2005: 448). Several strands of research have to be considered – namely literature on Europeanization as well as insights and approaches of studies dealing with cross-national policy convergence. Taking into account the non-obligatory and mainly intergovernmental character of the Bologna Process the main focus of the paper is on factors related to the effects of transnational communication. The inherent goal is to extend the research agenda on higher education (McLendon 2003: 184ff) and to leave behind the restriction of to analyse only a few cases by striving for a research design that allows for systematic testing and sufficient explanations of cross-national policy convergence at the interface between the Bologna Process and domestic factors.
The workshop presents ATHEN 1 (Annotation and Text Highlighting Environment), an extensible desktop-based annotation environment which supports more than just regular annotation. Besides being a general purpose annotation environment, ATHEN supports indexing and querying support of your data as well as the ability to automatically preprocess your data with Meta information. It is especially suited for those who want to extend existing general purpose annotation tools by implementing their own custom features, which cannot be fulfilled by other available annotation environments. On the according gitlab, we provide online tutorials, which demonstrate the use of specific features of ATHEN
Feedback utterances are among the most frequent in dialogue. Feedback is also a crucial aspect of all linguistic theories that take social interaction involving language into account. However, determining communicative functions is a notoriously difficult task both for human interpreters and systems. It involves an interpretative process that integrates various sources of information. Existing work on communicative function classification comes from either dialogue act tagging where it is generally coarse grained concerning the feed- back phenomena or it is token-based and does not address the variety of forms that feed- back utterances can take. This paper introduces an annotation framework, the dataset and the related annotation campaign (involving 7 raters to annotate nearly 6000 utterances). We present its evaluation not merely in terms of inter-rater agreement but also in terms of usability of the resulting reference dataset both from a linguistic research perspective and from a more applicative viewpoint.
In this paper, we investigate the practical applicability of Co-Training for the task of building a classifier for reference resolution. We are concerned with the question if Co-Training can significantly reduce the amount of manual labeling work and still produce a classifier with an acceptable performance.
Our paper describes an experiment aimed to assessment of lexical coverage in web corpora in comparison with the traditional ones for two closely related Slavic languages from the lexicographers’ perspective. The preliminary results show that web corpora should not be considered ― inferior, but rather ― different.
Common Crawl is a considerably large, heterogeneous multilingual corpus comprised of crawled documents from the internet, surpassing 20TB of data and distributed as a set of more than 50 thousand plain text files where each contains many documents written in a wide variety of languages. Even though each document has a metadata block associated to it, this data lacks any information about the language in which each document is written, making it extremely difficult to use Common Crawl for monolingual applications. We propose a general, highly parallel, multithreaded pipeline to clean and classify Common Crawl by language; we specifically design it so that it runs efficiently on medium to low resource infrastructures where I/O speeds are the main constraint. We develop the pipeline so that it can be easily reapplied to any kind of heterogeneous corpus and so that it can be parameterised to a wide range of infrastructures. We also distribute a 6.3TB version of Common Crawl, filtered, classified by language, shuffled at line level in order to avoid copyright issues, and ready to be used for NLP applications.
We present an implemented machine learning system for the automatic detection of nonreferential it in spoken dialog. The system builds on shallow features extracted from dialog transcripts. Our experiments indicate a level of performance that makes the system usable as a preprocessing filter for a coreference resolution system. We also report results of an annotation study dealing with the classification of it by naive subjects.
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.
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.
Germany’s diverse history in the 20th century raises the question of how social upheavals were constituted in and through political discourse. By analysing basic concepts, the research network “The 20th century in basic concepts” (based at the Leibniz institutes IDS, ZfL, ZZF) aims to identify continuities and discontinuities in political and social discourse. In this way, historical sediments of the present are to be uncovered and those challenges identified that emerged in the course of the 20th century and continue to shape political discourse until the present.
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.
Active learning has been applied to different NLP tasks, with the aim of limiting the amount of time and cost for human annotation. Most studies on active learning have only simulated the annotation scenario, using prelabelled gold standard data. We present the first active learning experiment for Word Sense Disambiguation with human annotators in a realistic environment, using fine-grained sense distinctions, and investigate whether AL can reduce annotation cost and boost classifier performance when applied to a real-world task.
This paper describes the application of probabilistic part of speech taggers to the Dzongkha language. A tag set containing 66 tags is designed, which is based on the Penn Treebank. A training corpus of 40,247 tokens is utilized to train the model. Using the lexicon extracted from the training corpus and lexicon from the available word list, we used two statistical taggers for comparison reasons. The best result achieved was 93.1% accuracy in a 10-fold cross validation on the training set. The winning tagger was thereafter applied to annotate a 570,247 token corpus.
The availability of large multi-parallel corpora offers an enormous wealth of material to contrastive corpus linguists, translators and language learners, if we can exploit the data properly. Necessary preparation steps include sentence and word alignment across multiple languages. Additionally, linguistic annotation such as partof- speech tagging, lemmatisation, chunking, and dependency parsing facilitate precise querying of linguistic properties and can be used to extend word alignment to sub-sentential groups. Such highly interconnected data is stored in a relational database to allow for efficient retrieval and linguistic data mining, which may include the statistics-based selection of good example sentences. The varying information needs of contrastive linguists require a flexible linguistic query language for ad hoc searches. Such queries in the format of generalised treebank query languages will be automatically translated into SQL queries.
This paper reports on an ongoing international project of compiling a freely accessible online Dictionary of German Loans in Polish Dialects. The dictionary will be the first comprehensive lexicographic compendium of its kind, serving as a complement to existing resources on German lexical loans in the literary or standard language. The empirical results obtained in the project will shed new light on the distribution of German loanwords among different dialects, also in comparison to the well-documented situation in written Polish. The dictionary will have a strong focus on the dialectal distribution of Polish dialectal variants for a given German etymon, accessible through interactive cartographic representations and corresponding search options. The editorial process is realized with dedicated collaborative web tools. The new resource will be published as an integrated part of an online information system for German lexical borrowings in other languages, the Lehnwortportal Deutsch, and is therefore highly cross-linked with other loanword dictionaries on Polish as well as Slavic and further European languages.
CLARIN contractual framework for sharing language data: the perspective of personal data protection
(2020)
The article analyses the responsibility for ensuring compliance with the General Data Protection Regulation (GDPR) in research settings. As a general rule, organisations are considered the data controller (responsible party for the GDPR compliance). Research constitutes a unique setting influenced by academic freedom. This raises the question of whether academics could be considered the controller as well. However, there are some court cases and policy documents on this issue. It is not settled yet. The analysis serves a preliminary analytical background for redesigning CLARIN contractual framework for sharing data.
Data Management is one of the core activities of all CLARIN centres providing data and services for the academia. In PARTHENOS, European initiatives and projects in the area of the humanities and social sciences assembled to compare policies and procedures. One of the areas of interest is data management. The data management landscape shows a lot of proliferation, for which an abstraction level is introduced to help centres, such as CLARIN centres, in the process of providing the best possible services to users with data management needs.
Since 2013 representatives of several French and German CMC corpus projects have developed three customizations of the TEI-P5 standard for text encoding in order to adapt the encoding schema and models provided by the TEI to the structural peculiarities of CMC discourse. Based on the three schema versions, a 4th version has been created which takes into account the experiences from encoding our corpora and which is specifically designed for the submission of a feature request to the TEI council. On our poster we would present the structure of this schema and its relations (commonalities and differences) to the previous schemas.
CMDI Explorer
(2021)
We present CMDI Explorer, a tool that empowers users to easily explore the contents of complex CMDI records and to process selected parts of them with little effort. The tool allows users, for instance, to analyse virtual collections represented by CMDI records, and to send collection items to other CLARIN services such as the Switchboard for subsequent processing. CMDI Explorer hence adds functionality that many users felt was lacking from the CLARIN tool space.
The paper’s purpose is to give an overview of the work on the Component Metadata Infrastructure (CMDI) that was implemented in the CLARIN research infrastructure. It explains, the underlying schema, the accompanying tools and services. It also describes the status and impact of the CMDI developments done within the CLARIN project and past and future collaborations with other projects.
This paper discusses how cognitive aspects can be incorporated into lexicographic meaning descriptions based on corpus-driven analysis. The new German Online dictionary “Paronyme − Dynamisch im Kontrast” is concerned with easily confused words such as effektiv/effizient, sensibel/sensitiv. It is currently in the process of being developed and it aims at adopting a more conceptual and encyclopedic approach to meaning. Contrastive entries emphasize usage, comparing conceptual categories and indicating the mapping of knowledge. Adaptable access to lexicographic details offers different perspectives on information, and authentic examples reflect prototypical structures.
Some of the cognitive features are demonstrated with the help of examples. Firstly, I will outline how patterns of usage imply conceptual categories as central ideas instead of sufficiently logical criteria of semantic distinction. In this way, linguistic findings correlate better with how users conceptualize language. Secondly, it is pointed out how collocates are family members and fillers in contexts. Thirdly, I will demonstrate how contextual structure and function are included by summarizing referential information. Details are drawn from corpus data; they are usage-based patterns illustrating conversational interaction and semantic negotiation in contemporary public discourse. Finally, I will show flexible consultation routines where the focus on structural knowledge changes.
Prominence has been widely studied on the word level and the syllable level. An extensive study comparing the two approaches is missing in the literature. This study investigates how word and syllable prominence relate to each other in German. We find that perceptual ratings based on the word level are more extreme than those based on the syllable level. The correlations between word prominence and acoustic features are greater than the correlations between syllable prominence and acoustic features.
We present an approach to making existing CLARIN web services usable for spoken language transcriptions. Our approach is based on a new TEI-based ISO standard for such transcriptions. We show how existing tool formats can be transformed to this standard, how an encoder/decoder pair for the TCF format enables users to feed this type of data through a WebLicht tool chain, and why and how web services operating directly on the standard format would be useful.
Converting and Representing Social Media Corpora into TEI: Schema and best practices from CLARIN-D
(2016)
The paper presents results from a curation project within CLARIN-D, in which an existing lMWord corpus of German chat communication has been integrated into the DEREKO and DWDS corpus infrastructures of the CLARIN-D centres at the Institute for the German Language (IDS, Mannheim) and at the Berlin-Brandenburg Academy of Sciences (BBAW, Berlin). The focus is on the solutions developed for converting and representing the corpus in a TEI format.
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 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.
This article reports on the on-going CoRoLa project, aiming at creating a reference corpus of contemporary Romanian (from 1945 onwards), opened for online free exploitation by researchers in linguistics and language processing, teachers of Romanian, students. We invest serious efforts in persuading large publishing houses and other owners of IPR on relevant language data to join us and contribute the project with selections of their text and speech repositories. The CoRoLa project is coordinated by two Computer Science institutes of the Romanian Academy, but enjoys cooperation of and consulting from professional linguists from other institutes of the Romanian Academy. We foresee a written component of the corpus of more than 500 million word forms, and a speech component of about 300 hours of recordings. The entire collection of texts (covering all functional styles of the language) will be pre-processed and annotated at several levels, and also documented with standardized metadata. The pre-processing includes cleaning the data and harmonising the diacritics, sentence splitting and tokenization. Annotation will include morpho-lexical tagging and lemmatization in the first stage, followed by syntactic, semantic and discourse annotation in a later stage.
Making research data publicly available for evaluation or reuse is a fundamental part of good scientific practice. However, regulations such as copyright law can prevent this practice and thereby hamper scientific progress. In Germany, text-based research disciplines have for a long time been mostly unable to publish corpora made from material outside of the public domain, effectively excluding contemporary works. While there are approaches to obfuscate text material in a way that it is no longer covered by the original copyright, many use cases still require the raw textual context for evaluation or follow-up research. Recent changes in copyright now permit text and data mining on copyrighted works. However, questions regarding reusability and sharing of such corpora at a later time are still not answered to a satisfying degree. We propose a workflow that allows interested third parties to access customized excerpts of protected corpora in accordance with current German copyright law and the soon to be implemented guidelines of the Digital Single Market directive. Our prototype is a very lightweight web interface that builds on commonly used repository software and web standards.
Corpus-based identification and disambiguation of reading indicators for German nominalizations
(2010)
Corpus data is often structurally and lexically ambiguous; corpus extraction methodologies thus must be made aware of ambiguities. Therefore, given an extraction task, all relevant ambiguities must be identified. To resolve these ambiguities, contextual data responsible for one or another reading is to be considered. In the context of our present work, German -ung-nominalizations and their sortal readings are under examination. A number of these nominalizations may be read as an event or a result, depending on the semantic group they belong to. Here, we concentrate on nominalizations of verbs of saying (henceforth: "verba dicendi"), identify their context partners and their influence on the sortal reading of the nominalizations in question. We present a tool which calculates the sortal reading of such nominalizations and thus may improve not only corpus extraction, but also e.g. machine translation. Lastly, we describe successful attempts to identify the correct sortal reading, conclusions and future work.
We present the use of count-based and predictive language models for exploring language use in the German Reference Corpus DeReKo. For collocation analysis along the syntagmatic axis we employ traditional association measures based on co-occurrence counts as well as predictive association measures derived from the output weights of skipgram word embeddings. For inspecting the semantic neighbourhood of words along the paradigmatic axis we visualize the high dimensional word embeddings in two dimensions using t-stochastic neighbourhood embeddings. Together, these visualizations provide a complementary, explorative approach to analysing very large corpora in addition to corpus querying. Moreover, we discuss count-based and predictive models w.r.t. scalability and maintainability in very large corpora.
Creating CorCenCC (Corpws Cenedlaethol Cymraeg Cyfoes - The National Corpus of Contemporary Welsh)
(2017)
CorCenCC is an interdisciplinary and multiinstitutional project that is creating a large-scale, open-source corpus of contemporary Welsh. CorCenCC will be the first ever large-scale corpus to represent spoken, written and electronicallymediated Welsh (compiling an initial data set of 10 million Welsh words), with a functional design informed, from the outset, by representatives of all anticipated academic and community user groups.
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.
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.
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.
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 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.
We discuss the modal uses of the Hausa exclusive particle sai (≈ only). We argue that the distribution of sai in modal environments provides evidence for the following claims on the composition of modal meaning that have been independently made in the literature: i) Future-oriented modality involves a prospective aspect operator that can be realized covertly in some languages (e.g. English, Kratzer 2012b) and overtly in others (e.g. Gitksan, Matthewson 2012, 2013). ii) Necessity interpretations arise from exhaustifying possibilities, i.e. an exhaustivity operator applying to existential modality (e.g. Kaufmann 2012 for the case of imperatives and Leffel 2012 for a relevant analysis of necessity meaning in Masalit). We show that future-oriented necessity in Hausa decomposes into EXH((PROSP)), with sai contributing exhaustivity.
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 present a language learning application that relies on grammars to model the learning outcome. Based on this concept we can provide a powerful framework for language learning exercises with an intuitive user interface and a high reliability. Currently the application aims to augment existing language classes and support students by improving the learner attitude and the general learning outcome. Extensions beyond that scope are promising and likely to be added in the future.
This paper describes general requirements for evaluating and documenting NLP tools with a focus on morphological analysers and the design of a Gold Standard. It is argued that any evaluation must be measurable and documentation thereof must be made accessible for any user of the tool. The documentation must be of a kind that it enables the user to compare different tools offering the same service, hence the descriptions must contain measurable values. A Gold Standard presents a vital part of any measurable evaluation process, therefore, the corpus-based design of a Gold Standard, its creation and problems that occur are reported upon here. Our project concentrates on SMOR, a morphological analyser for German that is to be offered as a web-service. We not only utilize this analyser for designing the Gold Standard, but also evaluate the tool itself at the same time. Note that the project is ongoing, therefore, we cannot present final results.
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.
This paper uses a devil’s advocate position to highlight the benefits of metadata creation for linguistic resources. It provides an overview of the required metadata infrastructure and shows that this infrastructure is in the meantime developed by various projects and hence can be deployed by those working with linguistic resources and archiving. Possible caveats of metadata creation are mentioned starting with user requirements and backgrounds, contribution to academic merits of researchers and standardisation. These are answered with existing technologies and procedures, referring to the Component Metadata Infrastructure (CMDI). CMDI provides an infrastructure and methods for adapting metadata to the requirements of specific classes of resources, using central registries for data categories, and metadata schemas. These registries allow for the definition of metadata schemas per resource type while reusing groups of data categories also used by other schemas. In summary, rules of best practice for the creation of metadata are given.
With an increasing amount of text data available it is possible to automatically extract a variety of information about language. One way to obtain knowledge about subtle relations and analogies between words is to observe words which are used in the same context. Recently, Mikolov et al. proposed a method to efficiently compute Euclidean word representations which seem to capture subtle relations and analogies between words in the English language. We demonstrate that this method also captures analogies in the German language. Furthermore, we show that we can transfer information extracted from large non-annotated corpora into small annotated corpora, which are then, in turn, used for training NLP systems.
To optimize the sharing and reuse of existing data, many funding organizations now require researchers to specify a management plan for research data. In such a plan, researchers are supposed to describe the entire life cycle of the research data they are going to produce, from data creation to formatting, interpretation, documentation, short-term storage, long-term archiving and data re-use. To support researchers with this task, we built DMPTY, a wizard that guides researchers through the essential aspects of managing data, elicits information from them, and finally, generates a document that can be further edited and linked to the original research proposal.
We investigate whether non-configurational languages, which display more word order variation than configurational ones, require more training data for a phenomenon to be parsed successfully. We perform a tightly controlled study comparing the dative alternation for English (a configurational language), German, and Russian (both non-configurational). More specifically, we compare the performance of a dependency parser when only canonical word order is present with its performance on data sets when all word orders are present. Our results show that for all languages, canonical data not only is easier to parse, but there exists no direct correspondence between the size of training sets containing free(er) word order variation and performance.
Sentiment analysis has so far focused on the detection of explicit opinions. However, of late implicit opinions have received broader attention, the key idea being that the evaluation of an event type by a speaker depends on how the participants in the event are valued and how the event itself affects the participants. We present an annotation scheme for adding relevant information, couched in terms of so-called effect functors, to German lexical items. Our scheme synthesizes and extends previous proposals. We report on an inter-annotator agreement study. We also present results of a crowdsourcing experiment to test the utility of some known and some new functors for opinion inference where, unlike in previous work, subjects are asked to reason from event evaluation to participant evaluation.
The Component MetaData Infrastructure (CMDI) is the dominant framework for describing language resources according to ISO 24622 (ISO/TC 37/SC 4, 2015). Within the CLARIN world, CMDI has become a huge success. The Virtual Language Observatory (VLO) now holds over 800.000 resources, all described with CMDI-based metadata. With the metadata being harvested from about thirty centres, there is a considerable amount of heterogeneity in the data. In part, there is some use of controlled vocabularies to keep data heterogeneity in check, say when describing the type of a resource, or the country the resource is originating from. However, when CMDI data refers to the names of persons or organisations, strings are used in a rather uncontrolled manner. Here, the CMDI community can learn from libraries and archives who maintain standardised lists for all kinds of names. In this paper, we advocate the use of freely available authority files that support the unique identification of persons, organisations, and more. The systematic use of authority records enhances the quality of the metadata, hence improves the faceted browsing experience in the VLO, and also prepares the sharing of CMDI-based metadata with the data in library catalogues.
We evaluate a graph-based dependency parser on DeReKo, a large corpus of contemporary German. The dependency parser is trained on the German dataset from the SPMRL 2014 Shared Task which contains text from the news domain, whereas DeReKo also covers other domains including fiction, science, and technology. To avoid the need for costly manual annotation of the corpus, we use the parser’s probability estimates for unlabeled and labeled attachment as main evaluation criterion. We show that these probability estimates are highly correlated with the actual attachment scores on a manually annotated test set. On this basis, we compare estimated parsing scores for the individual domains in DeReKo, and show that the scores decrease with increasing distance of a domain to the training corpus.
This paper presents the QUEST project and describes concepts and tools that are being developed within its framework. The goal of the project is to establish quality criteria and curation criteria for annotated audiovisual language data. Building on existing resources developed by the participating institutions earlier, QUEST develops tools that could be used to facilitate and verify adherence to these criteria. An important focus of the project is making these tools accessible for researchers without substantial technical background and helping them produce high-quality data. The main tools we intend to provide are the depositors’ questionnaire and automatic quality assurance, both developed as web applications. They are accompanied by a Knowledge base, which will contain recommendations and descriptions of best practices established in the course of the project. Conceptually, we split linguistic data into three resource classes (data deposits, collections and corpora). The class of a resource defines the strictness of the quality assurance it should undergo. This division is introduced so that too strict quality criteria do not prevent researchers from depositing their data.
This paper presents the QUEST project and describes concepts and tools that are being developed within its framework. The goal of the project is to establish quality criteria and curation criteria for annotated audiovisual language data. Building on existing resources developed by the participating institutions earlier, QUEST also develops tools that could be used to facilitate and verify adherence to these criteria. An important focus of the project is making these tools accessible for researchers without substantial technical background and helping them produce high-quality data. The main tools we intend to provide are a questionnaire and automatic quality assurance for depositors of language resources, both developed as web applications. They are accompanied by a knowledge base, which will contain recommendations and descriptions of best practices established in the course of the project. Conceptually, we consider three main data maturity levels in order to decide on a suitable level of strictness of the quality assurance. This division has been introduced to avoid that a set of ideal quality criteria prevent researchers from depositing or even assessing their (legacy) data. The tools described in the paper are work in progress and are expected to be released by the end of the QUEST project in 2022.
Streefkerk defines prominence as the perceptually outstanding parts in spoken language. An optimal rating scale for syllable prominence has not been found yet. This paper evaluates a 4-point, an 11-point, a 31-point, and a continuous scale for the rating of syllable prominence and gives support for scales using a higher number of levels. Priming effects found by Arnold, et al., could only be replicated using the 31-point scale.
To improve grammatical function labelling for German, we augment the labelling component of a neural dependency parser with a decision history. We present different ways to encode the history, using different LSTM architectures, and show that our models yield significant improvements, resulting in a LAS for German that is close to the best result from the SPMRL 2014 shared task (without the reranker).
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.
Unknown words are a challenge for any NLP task, including sentiment analysis. Here, we evaluate the extent to which sentiment polarity of complex words can be predicted based on their morphological make-up. We do this on German as it has very productive processes of derivation and compounding and many German hapax words, which are likely to bear sentiment, are morphologically complex. We present results of supervised classification experiments on new datasets with morphological parses and polarity annotations.
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 collection of (currently) about 5.500 commands directed to voice-controlled virtual assistants (VAs) by sixteen initial users of a VA system in their homes. The collection comprises recordings captured by the VA itself and with a conditional voice recorder (CVR) selectively capturing recordings including the VA-directed commands plus some surrounding context. Next to a description of the collection, we present initial findings on the patterns of use of the VA systems during the first weeks after installation, including usage timing, the development of usage frequency, distributions of sentence structures across commands, and (the development of) command success rates. We discuss the advantages and disadvantages of the applied collection-specific recording approach and describe potential research questions that can be investigated in the future, based on the collection, as well as the merit of combining quantitative corpus linguistic approaches with qualitative in-depth analyses of single cases.
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.
The classification of verbs in Levin's (1993) English Verb Classes and Alternations: A preliminary Investigation, on the basis of both intuitive semantic grouping and their participation in valence alternations, is often used by the NLP community as evidence of the semantic similarity of verbs (Jing & McKeown 1998; Lapata & Brew 1999; Kohl et al. 1998). In this paper, we compare the Levin classification with the work of the FrameNet project (Fillmore & Baker 2001), where words (not just verbs) are grouped according to the conceptual structures (frames) that underlie them and their combinatorial patterns are inductively derived from corpus evidence. This means that verbs grouped together in FrameNet (FN) might be semantically similar but have different (or no) alternations, and that verbs which share the same alternation might be represented in two different semantic frames.
In 2010, ISO published a standard for syntactic annotation, ISO 24615:2010 (SynAF). Back then, the document specified a comprehensive reference model for the representation of syntactic annotations, but no accompanying XML serialisation. ISO’s subcommittee on language resource management (ISO TC 37/SC 4) is working on making the SynAF serialisation ISOTiger an additional part of the standard. This contribution addresses the current state of development of ISOTiger, along with a number of open issues on which we are seeking community feedback in order to ensure that ISOTiger becomes a useful extension to the SynAF reference model.
This paper outlines the broad research context and rationale for a new international comparable corpus (ICC). The ICC is to be largely modelled on the text categories and their quantities the International Corpus of English with only a few changes. The corpus will initially begin with nine European languages but others may join in due course. The paper reports on those and other agreements made at the inaugural planning meeting in Prague on 22-23 June 2017. It also sets out the project’s goals for its first two years.