Refine
Year of publication
Document Type
- Part of a Book (249)
- Article (200)
- Conference Proceeding (150)
- Book (39)
- Working Paper (12)
- Other (11)
- Part of Periodical (10)
- Preprint (10)
- Doctoral Thesis (2)
- Course Material (1)
Language
- English (687) (remove)
Is part of the Bibliography
- yes (687) (remove)
Keywords
- Korpus <Linguistik> (203)
- Deutsch (185)
- Interaktion (74)
- Konversationsanalyse (69)
- Gesprochene Sprache (52)
- Computerlinguistik (43)
- Forschungsdaten (40)
- Annotation (38)
- Englisch (34)
- Wörterbuch (34)
Publicationstate
- Veröffentlichungsversion (402)
- Zweitveröffentlichung (136)
- Postprint (88)
- Ahead of Print (5)
- Preprint (3)
- Erstveröffentlichung (1)
Reviewstate
- Peer-Review (431)
- (Verlags)-Lektorat (107)
- Peer-review (10)
- Peer-Revied (5)
- Verlags-Lektorat (5)
- Qualifikationsarbeit (Dissertation, Habilitationsschrift) (2)
- (Verlags-)Lektorat (1)
- (Verlags-)lektorat (1)
- Abschlussarbeit (Bachelor, Master, Diplom, Magister) (Bachelor, Master, Diss.) (1)
- Peer review (1)
Publisher
This paper reports on the efforts of twelve national teams in building the International Comparable Corpus (ICC; https://korpus.cz/icc) that will contain highly comparable datasets of spoken, written and electronic registers. The languages currently covered are Czech, Finnish, French, German, Irish, Italian, Norwegian, Polish, Slovak, Swedish and, more recently, Chinese, as well as English, which is considered to be the pivot language. The goal of the project is to provide much-needed data for contrastive corpus-based linguistics. The ICC corpus is committed to the idea of re-using existing multilingual resources as much as possible and the design is modelled, with various adjustments, on the International Corpus of English (ICE). As such, ICC will contain approximately the same balance of forty percent of written language and 60 percent of spoken language distributed across 27 different text types and contexts. A number of issues encountered by the project teams are discussed, ranging from copyright and data sustainability to technical advances in data distribution.
In informal interaction, speakers rarely thank a person who has complied with a request. Examining data from British English, German, Italian, Polish, and Telugu, we ask when speakers do thank after compliance. The results show that thanking treats the other’s assistance as going beyond what could be taken for granted in the circumstances. Coupled with the rareness of thanking after requests, this suggests that cooperation is to a great extent governed by expectations of helpfulness, which can be long-standing, or built over the course of a particular interaction. The higher frequency of thanking in some languages (such as English or Italian) suggests that cultures differ in the importance they place on recognizing the other’s agency in doing as requested.
Drawing on research from conversation analysis and developmental psychology, we point to the existence of “supporters” of morally responsible agency in everyday interaction: causes of our behavior that we are often unaware of, but that would make goodenough reasons for our actions, were we made aware of them.
Sometimes in interaction, a speaker articulates an overt interpretation of prior talk. Such moments have been studied as involving the repair of a problem with the other’s talk or as formulating an understanding of the matter at hand. Stepping back from the established notions of formulations and repair, we examine the variety of actions speakers do with the practice of offering an interpretation, and the order within this domain. Results show half a dozen usage types of interpretations in mundane interaction. These form a largely continuous territory of action, with recognizably distinct usage types as well as cases falling between these (proto)typical uses. We locate order in the domain of interpretations using the method of semantic maps and show that, contrary to earlier assumptions in the literature, interpretations that formulate an understanding of the matter at hand are actually quite pervasive in ordinary talk. These findings contribute to research on action formation and advance our understanding of understanding in interaction. Data are video- and audio-recordings of mundane social interaction in the German language from a variety of settings.
The present paper explores how rules are enforced and talked about in everyday life. Drawing on a corpus of board game recordings across European languages, we identify a sequential and praxeological context for rule talk. After a game rule is breached, a participant enforces proper play and then formulates a rule with an impersonal deontic statement (e.g. “It’s not allowed to do this”). Impersonal deontic statements express what may or may not be done without tying the obligation to a particular individual. Our analysis shows that such statements are used as part of multi-unit and multi-modal turns where rule talk is accomplished through both grammatical and embodied means. Impersonal deontic statements serve multiple interactional goals: they account for having changed another’s behavior in the moment and at the same time impart knowledge for the future. We refer to this complex action as an “instruction.” The results of this study advance our understanding of rules and rule-following in everyday life, and of how resources of language and the body are combined to enforce and formulate rules.
We examine moments in social interaction in which a person formulates what another thinks or believes. Such formulations of belief constitute a practice with specifiable contexts and consequences. Belief formulations treat aspects of the other person's prior conduct as accountable on the basis that it provided a new angle on a topic, or otherwise made a surprising contribution within an ongoing course of actions. The practice of belief formulations subjectivizes the content that the other articulated and thereby topicalizes it, mobilizing commitment to that position, an account, or further elaboration. We describe how the practice can be put to work in different activity contexts: sometimes it is designed to undermine the other's position as a subjective 'mere belief', at other times it serves to mobilize further topic talk. Throughout, belief formulations show themselves to be a method by which we get to know ourselves and each other as mental agents.
In the management of cooperation, the fit of a requested action with what the addressee is presently doing is a pervasively relevant consideration. We present evidence that imperative turns are adapted to, and reflexively create, contexts in which the other person is committed to the course of action advanced by the imperative. This evidence comes from systematic variation in the design of imperative turns, relative to the fittedness of the imperatively mandated action to the addressee’s ongoing trajectory of actions, what we call the “dine of commitment”. We present four points on this dine: Responsive imperatives perform an operation on the deontic dimension of what the addressee has announced or already begun to do (in particular its permissibility); local-project imperatives formulate a new action advancing a course of action in which the addressee is already actively engaged; global-project-imperatives target a next task for which the addressee is available on the grounds of their participation in the overall event, and in the absence of any competing work; and competitive imperatives draw on a presently otherwise engaged addressee on the grounds of their social commitment to the relevant course of actions. These four turn shapes are increasingly complex, reflecting the interactional work required to bridge the increasing distance between what the addressee is currently doing, and what the imperative mandates. We present data from German and Polish informal and institutional settings.
Linguistic relativists have traditionally asked 'how language influences thought', but conversation analysts and anthropological linguists have moved the focus from thought to social action. We argue that 'social action' should in this context not become simply a new dependent variable, because the formulation 'does language influence action' suggests that social action would already be meaningfully constituted prior to its local (verbal and multi-modal) accomplishment. We draw on work by the gestalt psychologist Karl Duncker to show that close attention to action-in-a-situation helps us ground empirical work on cross-cultural diversity in an appreciation of the invariances that make culture-specific elements of practice meaningful.
This article makes an empirical and a methodological contribution to the comparative study of action. The empirical contribution is a comparative study of three distinct types of action regularly accomplished with the turn format du meinst x (“you mean/think x”) in German: candidate understandings, formulations of the other’s mind, and requests for a judgment. These empirical materials are the basis for a methodological exploration of different levels of researcher abstraction in the comparative study of action. Two levels are examined: the (coarser) level of conditionally relevant responses (what a response speaker must do to align with the action of the prior turn) and the (finer) level of “full alignment” (what a response speaker can do to align with the action of a prior turn). Both levels of abstraction provide empirically viable and analytically interesting descriptive concepts for the comparative study of action. Data are in German.
This article makes an empirical and a methodological contribution to the comparative study of action. The empirical contribution is a comparative study of three distinct types of action regularly accomplished with the turn format du meinst x (“you mean/think x”) in German: candidate understandings, formulations of the other’s mind, and requests for a judgment. These empirical materials are the basis for a methodological exploration of different levels of researcher abstraction in the comparative study of action. Two levels are examined: the (coarser) level of conditionally relevant responses (what a response speaker must do to align with the action of the prior turn) and the (finer) level of “full alignment” (what a response speaker can do to align with the action of a prior turn). Both levels of abstraction provide empirically viable and analytically interesting descriptive concepts for the comparative study of action. Data are in German.
This chapter describes the resources that speakers of Polish use when recruiting assistance and collaboration from others in everyday social interaction. The chapter draws on data from video recordings of informal conversation in Polish, and reports language-specific findings generated within a large-scale comparative project involving eight languages from five continents (see other chapters of this volume). The resources for recruitment described in this chapter include linguistic structures from across the levels of grammatical organization, as well as gestural and other visible and contextual resources of relevance to the interpretation of action in interaction. The presentation of categories of recruitment, and elements of recruitment sequences, follows the coding scheme used in the comparative project (see Chapter 2 of the volume). This chapter extends our knowledge of the structure and usage of Polish with detailed attention to the properties of sequential structure in conversational interaction. The chapter is a contribution to an emerging field of pragmatic typology.
When formulating a request for an object, speakers can choose among different grammatical resources that would all serve the overall purpose. This paper examines the social contexts indexed and created by the choice of the turn format can I have x to request a shared good (the pepper grinder, a tissue from a box on the table, etc.) in British English informal interaction. The analysis is based on a video corpus of approximately 25 h of everyday interaction among family and friends. In its home environment, a request in the format can I have x treats the other as being in control over the relevant material object, a control that is the contingent outcome of ongoing courses of action. This contingent control over a shared good produces an obligation to make it available. This analysis is supported by an examination of similarly formatted request turns in other languages, of can I have x in another interactional environment (after a relevant offer has been made) in British English, and of deviant cases. The results highlight the intimate connection of request format selection to the present engagements of (prospective) request recipients.
Temporal frames of reference
(2010)
‘Can’ and ‘must’-type modal verbs in the direct sanctioning of misconduct across European languages
(2023)
Deontic meanings of obligation and permissibility have mostly been studied in relation to modal verbs, even though researchers are aware that such meanings can be conveyed in other ways (consider, for example, the contributions to Nuyts/van der Auwera (eds.) 2016). This presentation reports on an ongoing project that examines deontic meaning but takes as its starting point not a type of linguistic structure but a particular kind of social moment that presumably attracts deontic talk: The management of potentially ‚unacceptable‘ or untoward actions (taking the last bread roll at breakfast, making a disallowed move during a board game, etc.). Data come from a multi-language parallel video corpus of everyday social interaction in English, German, Italian, and Polish. Here, we focus on moments in which one person sanctions another’s behavior as unacceptable. Using interactional-linguistic methods (Couper-Kuhlen/Selting 2018), we examine similarities and differences across these four languages in the use of modal verbs as part of such sanctioning attempts. First results suggest that modal verbs are not as common in the sanctioning of misconduct as one might expect. Across the four languages, only between 10%–20% of relevant sequences involve a modal verb. Most of the time, in this context, speakers achieve deontic meaning in other ways (e.g., infinitives such as German nicht so schmatzen, ‚no smacking‘). This raises the question what exactly modal verbs, on those relatively rare occasions when they are used, contribute to the accomplishment of deontic meaning. The reported study pursues this question in two ways: 1) By considering similarities across languages in the ways that modal verbs interact with other (verbal) means in the sanctioning of misconduct.; 2) By considering differences across languages in the use of modal verbs. Here, we find that the relevant modal verbs are used similarly in some activity contexts (enforcing rules during board games), but less so in other activity contexts (mundane situations with no codified rules). In sum, the presented study adds to cross-linguistically grounded knowledge about deontic meaning and its relationships to linguistics structures.
A constructicon, i.e., a structured inventory of constructions, essentially aims at documenting functions of lexical and grammatical constructions. Among other parameters, so-called constructional collo-profiles, as introduced by Herbst (2018, 2020), are conclusive for determining constructional meanings. They provide information on how relevant individual words are for construction slots, they hint at usage preferences of constructions and serve as a helpful indicator for semantic peculiarities of constructions. However, even though collo-profiles constitute an indispensable component of constructicon entries, they pose major challengers for constructicographers: For a constructicographic enterprise it is not feasible to conduct collostructional analyses for hundreds or even thousands of constructions. In this article, we introduce a procedure based on the large language model BERT that allows to predict collo-profiles without having to extensively annotate instances of constructions in a given corpus. Specifically, by discussing the constructions X macht Y ADJP (‘x makes Y ADJ’, e.g. he drives him crazy) and N1 PREP N1 (e.g., bumper to bumper, constructions over constructions), we show how the developed automated system generates collo-profiles based on a limited number of annotated instances. Finally, we place collo-profiles alongside other dimensions of constructional meanings included in the German Constructicon.
Speakers’ linguistic experience is for the most part experience with language as used in conversational interaction. Though highly relevant for usage-based linguistics, the study of such data is as yet often left to other frameworks such as conversation analysis and interactional linguistics (Couper-Kuhlen and Selting 2001). On the basis of a case study of salient usage patterns of the two German motion verbs kommen and gehen in spontaneous conversation, the present paper argues for a methodological integration of quantitative corpus-linguistic methods with qualitative conversation analytic approaches to further the usage-based study of conversational interaction.
Novel formats of construction-based description hold great potential for phenomena that fall through the cracks in traditional kinds of linguistic reference works. On the example of German verb argument structure constructions with a prepositional object, we demonstrate that a construction-based description of such phenomena is superior to existing lexicographic and grammaticographic treatments, but that it also poses a number of new problems. The most fundamental of these relates to the fact that construction-based analyses can be proposed on different levels of abstraction. We illustrate pertinent problems relating to the precise identification of constructional form and meaning and suggest a multi-layered descriptive format for web-based electronic reference constructica that can accommodate these challenges. Semantically, the proposed solution integrates both lumping and splitting perspectives on constructional grain size and permits users to flexibly zoom in and out on individual elements in the resource. Formally, it can capture variation in the number and marking of realised arguments as found in e.g. passives and transitivity alternations. Aspects of the theoretical controversy between Construction Grammar and Valency Theory are addressed where relevant, but our focus is on questions of description and the practical implementation of construction-based analyses in a suitable type of linguistic reference work.
Localism
(2017)
Localist hypothesis
(2017)
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.
Looking at gestures as a means for communication, they can serve conversational participants at several levels. As co-speech gestures, they can add information to the verbally expressed content and they can serve to manage turn-taking. In order to look closer at the interplay between these resources in face-to face conversation, we annotated hand gestures, syntactic completion points and the related turn-organisation, and measured the timing of gesture strokes and their lexical/phrasal referent. In a case study on German, we observe the trend that speakers vary less in gesturelexis on- and offsets when keeping the turn after syntactic completions than at speaker changes, backchannel or other locations of a conversation. This indicates that timing properties of non-verbal cues interact with verbal cues to manage turn-taking.
Many studies on dictionary use presuppose that users do indeed consult lexicographic resources. However, little is known about what users actually do when they try to solve language problems on their own. We present an observation study where learners of German were allowed to browse the web freely while correcting erroneous German sentences. In this paper, we are focusing on the multi-methodological approach of the study, especially the interplay between quantitative and qualitative approaches. In one example study, we will show how the analysis of verbal protocols, the correction task and the screen recordings can reveal the effects of intuition, language (learning) awareness, and determination on the accuracy of the corrections. In another example study, we will show how preconceived hypotheses about the problem at hand might hinder participants from arriving at the correct solution.
Dictionaries have been part and parcel of literate societies for many centuries. They assist in communication, particularly across different languages, to aid in understanding, creating, and translating texts. Communication problems arise whenever a native speaker of one language comes into contact with a speaker of another language. At the same time, English has established itself as a lingua franca of international communication. This marked tendency gives lexicography of English a particular significance, as English dictionaries are used intensively and extensively by huge numbers of people worldwide.
We present ESDexplorer (https://owid.shinyapps.io/ESDexplorer), a browser application which allows the user to explore the data from a large European survey on dictionary use and culture. We built ESDexplorer with several target groups in mind: our cooperation partners, other researchers, and a more general public interested in the results. Also, we present in detail the architecture and technological realisation of the application and discuss some legal aspects of data protection that motivated some architectural choices.
The coronavirus pandemic may be the largest crisis the world has had to face since World War II. It does not come as a surprise that it is also having an impact on language as our primary communication tool. In this short paper, we present three inter-connected resources that are designed to capture and illustrate these effects on a subset of the German language: An RSS corpus of German-language newsfeeds (with freely available untruncated frequency lists), a continuously updated HTML page tracking the diversity of the vocabulary in the RSS corpus and a Shiny web application that enables other researchers and the broader public to explore the corpus in terms of basic frequencies.
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.
We introduce DeReKoGram, a novel frequency dataset containing lemma and part-of-speech (POS) information for 1-, 2-, and 3-grams from the German Reference Corpus. The dataset contains information based on a corpus of 43.2 billion tokens and is divided into 16 parts based on 16 corpus folds. We describe how the dataset was created and structured. By evaluating the distribution over the 16 folds, we show that it is possible to work with a subset of the folds in many use cases (e.g., to save computational resources). In a case study, we investigate the growth of vocabulary (as well as the number of hapax legomena) as an increasing number of folds are included in the analysis. We cross-combine this with the various cleaning stages of the dataset. We also give some guidance in the form of Python, R, and Stata markdown scripts on how to work with the resource.
Neologisms, i.e., new words or meanings, are finding their way into everyday language use all the time. In the process, already existing elements of a language are recombined or linguistic material from other languages is borrowed. But are borrowed neologisms accepted similarly well by the speech community as neologisms that were formed from “native” material? We investigate this question based on neologisms in German. Building on the corresponding results of a corpus study, we test the hypothesis of whether “native” neologisms are more readily accepted than those borrowed from English. To do so, we use a psycholinguistic experimental paradigm that allows us to estimate the degree of uncertainty of the participants based on the mouse trajectories of their responses. Unexpectedly, our results suggest that the neologisms borrowed from English are accepted more frequently, more quickly, and more easily than the “native” ones. These effects, however, are restricted to people born after 1980, the so-called millenials. We propose potential explanations for this mismatch between corpus results and experimental data and argue, among other things, for a reinterpretation of previous corpus studies.
We present an empirical study addressing the question whether, and to which extent, lexicographic writing aids improve text revision results. German university students were asked to optimise two German texts using (1) no aids at all, (2) highlighted problems, or (3) highlighted problems accompanied by lexicographic resources that could be used to solve the specific problems. We found that participants from the third group corrected the largest number of problems and introduced the fewest semantic distortions during revision. Also, they reached the highest overall score and were most efficient (as measured in points per time). The second group with highlighted problems lies between the two other groups in almost every measure we analysed. We discuss these findings in the scope of intelligent writing environments, the effectiveness of writing aids in practical usage situations and teaching dictionary skills.
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.
This replication study aims to investigate a potential bias toward addition in the German language, building upon previous findings of Winter and colleagues who identified a similar bias in English. Our results confirm a bias in word frequencies and binomial expressions, aligning with these previous findings. However, the analysis of distributional semantics based on word vectors did not yield consistent results for German. Furthermore, our study emphasizes the crucial role of selecting appropriate translational equivalents, highlighting the significance of considering language-specific factors when testing for such biases for languages other than English.
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.
The actual or anticipated impact of research projects can be documented in scientific publications and project reports. While project reports are available at varying level of accessibility, they might be rarely used or shared outside of academia. Moreover, a connection between outcomes of actual research project and potential secondary use might not be explicated in a project report. This paper outlines two methods for classifying and extracting the impact of publicly funded research projects. The first method is concerned with identifying impact categories and assigning these categories to research projects and their reports by extension by using subject matter experts; not considering the content of research reports. This process resulted in a classification schema that we describe in this paper. With the second method which is still work in progress, impact categories are extracted from the actual text data.
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 present an approach for modeling German negation in open-domain fine grained sentiment analysis. Unlike most previous work in sentiment analysis, we assume that negation can be conveyed by many lexical units (and not only common negation words) and that different negation words have different scopes. Our approach is examined on a new dataset comprising sentences with mentions of polar expressions and various negation words. We identify different types of negation words that have the same scopes. We show that already negation modeling based on these types largely outperforms traditional negation models which assume the same scope for all negation words and which employ a window-based scope detection rather than a scope detection based on syntactic information.
We present the pilot edition of the GermEval Shared Task on the Identification of Offensive Language. This shared task deals with the classification of German tweets from Twitter. It comprises two tasks, a coarse-grained binary classification task and a fine-grained multi-class classification task. The shared task had 20 participants submitting 51 runs for the coarse-grained task and 25 runs for the fine-grained task. Since this is a pilot task, we describe the process of extracting the raw-data for the data collection and the annotation schema. We evaluate the results of the systems submitted to the shared task. The shared task homepage can be found at https://projects.cai. fbi.h-da.de/iggsa/
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.
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 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.
Negation is an important contextual phenomenon that needs to be addressed in sentiment analysis. Next to common negation function words, such as not or none, there is also a considerably large class of negation content words, also referred to as shifters, such as the verbs diminish, reduce or reverse. However, many of these shifters are ambiguous. For instance, spoil as in spoil your chance reverses the polarity of the positive polar expression chance while in spoil your loved ones, no negation takes place. We present a supervised learning approach to disambiguating verbal shifters. Our approach takes into consideration various features, particularly generalization features.
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.