We present the DURel tool implementing the annotation of semantic proximity between word uses into an online, open source interface. The tool supports standardized human annotation as well as computational annotation, building on recent advances with Word-in-Context models. Annotator judgments are clustered with automatic graph clustering techniques and visualized for analysis. This allows to measure word senses with simple and intuitive micro-task judgments between use pairs, requiring minimal preparation efforts. The tool offers additional functionalities to compare the agreement between annotators to guarantee the inter-subjectivity of the obtained judgments and to calculate summary statistics over the annotated data giving insights into sense frequency distributions, semantic variation or changes of senses over time.

Predicting the compositionality of noun compounds such as climate change and tennis elbow is a vital component in natural language understanding. While most previous computational methods that automatically determine the semantic relatedness between compounds and their constituents have applied a synchronic perspective, the current study investigates what diachronic changes in contexts and semantic topics of compounds and constituents reveal about the compounds’ present-day degrees of compositionality. We define a binary classification task that utilizes two diachronic vector spaces based on contextual co-occurrences and semantic topics, and demonstrate that diachronic changes in cosine similarities – measured over context or topic distributions – uncover patterns that distinguish between compounds with low and high present-day compositionality. Despite fewer dimensions in the topic models, the topic space performs on par with the co-occurrence space and captures rather similar information. Temporal similarities between compounds and modifiers as well as between compounds and their prepositional paraphrases predict the compounds’ present-day compositionality with accuracy >0.7.

We present a comprehensive computational study of the under-investigated phenomenon of personal name compounds (PNCs) in German such as Willkommens-Merkel (‘Welcome-Merkel’). Prevalent in news, social media, and political discourse, PNCs are hypothesized to exhibit an evaluative function that is reflected in a more positive or negative perception as compared to the respective personal full name (such as Angela Merkel). We model 321 PNCs and their corresponding full names at discourse level, and show that PNCs bear an evaluative nature that can be captured through a variety of computational methods. Specifically, we assess through valence information whether a PNC is more positively or negatively evaluative than the person’s name, by applying and comparing two approaches using (i) valence norms and (ii) pre-trained language models (PLMs). We further enrich our data with personal, domain-specific, and extra-linguistic information and perform a range of regression analyses revealing that factors including compound and modifier valence, domain, and political party membership influence how a PNC is evaluated.

Multiword expressions (MWEs) are composed of multiple words and exhibit variable degrees of compositionality. As such, their meanings are notoriously difficult to model, and it is unclear to what extent this issue affects transformer architectures. Addressing this gap, we provide the first in-depth survey of MWE processing with transformer models. We overall find that they capture MWE semantics inconsistently, as shown by reliance on surface patterns and memorized information. MWE meaning is also strongly localized, predominantly in early layers of the architecture. Representations benefit from specific linguistic properties, such as lower semantic idiosyncrasy and ambiguity of target expressions. Our findings overall question the ability of transformer models to robustly capture fine-grained semantics. Furthermore, we highlight the need for more directly comparable evaluation setups.

With the advent of diffusion-based image generation models such as DALL-E, Stable Diffusion and Midjourney, high quality images can be easily generated using textual inputs. It is unclear, however, to what extent the generated images resemble human mental representations, especially regarding abstract event knowledge. We analyse the capability of four state-of-the-art models in generating images of verb-object event pairs when we systematically manipulate the degrees of abstractness of both the verbs and the object nouns. Human judgements assess the generated images and demonstrate that DALL-E is strongest for event pairs with concrete nouns (e.g., “pour water”; “believe person”), while Midjourney is preferred for event pairs with abstract nouns (e.g., “raise awareness”; “remain mystery”), irrespective of the concreteness of the verb. Across models, humans were most unsatisfied with images of events pairs that combined concrete verbs with abstract direct-object nouns (e.g., “speak truth”), and an additional ad-hoc annotation contributes this to its potential for figurative language.

Research on metaphor detection (MD) in a multilingual setup has recently gained momentum. As for many tasks, it is however unclear how the amount of data used to pretrain large language models affects the performance, and whether non-neural models might provide a reasonable alternative, especially for MD in low-resource languages. This paper compares neural and non-neural cross-lingual models for English as the source language and Russian, German and Latin as target languages. In a series of experiments we show that the neural cross-lingual adapter architecture MAD-X performs best across target languages. Zero-shot classification with mBERT achieves decent results above the majority baseline, while few-shot classification with mBERT heavily depends on shot-selection, which is inconvenient in a cross-lingual setup where no validation data for the target language exists. The non-neural model, a random forest classifier with conceptual features, is outperformed by the neural models. Overall, we recommend MAD-X for metaphor detection not only in high-resource but also in low-resource scenarios regarding the amounts of pretraining data for mBERT.

The interplay of cultural and linguistic elements that characterizes metaphorical language poses a substantial challenge for both human comprehension and machine processing. This challenge goes beyond monolingual settings and becomes particularly complex in translation, even more so in automatic translation. We present VOLIMET, a corpus of 2,916 parallel sentences containing gold standard alignments of metaphorical verb-object pairs and their literal paraphrases, e.g., tackle/address question, from English to German and French. On the one hand, the parallel nature of our corpus enables us to explore monolingual patterns for metaphorical vs. literal uses in English. On the other hand, we investigate different aspects of cross-lingual translations into German and French and the extent to which metaphoricity and literalness in the source language are transferred to the target languages. Monolingually, our findings reveal clear preferences in using metaphorical or literal uses of verb-object pairs. Cross-lingually, we observe a rich variability in translations as well as different behaviors for our two target languages.

This study investigates the performance of SigLIP, a state-of-the-art Vision-Language Model (VLM), in predicting labels for images depicting 1,278 concepts. Our analysis across 300 images per concept shows that the model frequently predicts the exact user-tagged labels, but similarly, it often predicts labels that are semantically related to the exact labels in various ways: synonyms, hypernyms, co-hyponyms, and associated words, particularly for abstract concepts. We then zoom into the diversity of the user tags of images and word associations for abstract versus concrete concepts. Surprisingly, not only abstract but also concrete concepts exhibit significant variability, thus challenging the traditional view that representations of concrete concepts are less diverse.

We present a multi-task learning approach to predicting semantic plausibility by leveraging 50+ adapters categorized into 17 tasks within an efficient training framework. Across four plausibility datasets in English of varying size and linguistic constructions, we compare how models provided with knowledge from a range of NLP tasks perform in contrast to models without external information. Our results show that plausibility prediction benefits from complementary knowledge (e.g., provided by syntactic tasks) are significant but non-substantial, while performance may be hurt when injecting knowledge from an unsuitable task. Similarly important, we find that knowledge transfer may be hindered by class imbalance, and demonstrate the positive yet minor effect of balancing training data, even at the expense of size.

We propose a novel approach to learn domain-specific plausible materials for components in the vehicle repair domain by probing Pretrained Language Models (PLMs) in a cloze task style setting to overcome the lack of annotated datasets. We devise a new method to aggregate salient predictions from a set of cloze query templates and show that domain-adaptation using either a small, high-quality or a customized Wikipedia corpus boosts performance. When exploring resource-lean alternatives, we find a distilled PLM clearly outperforming a classic pattern-based algorithm. Further, given that 98% of our domain-specific components are multiword expressions, we successfully exploit the compositionality assumption as a way to address data sparsity.

To date, transformer-based models such as BERT have been less successful in predicting compositionality of noun compounds than static word embeddings. This is likely related to a suboptimal use of the encoded information, reflecting an incomplete grasp of how the models represent the meanings of complex linguistic structures. This paper investigates variants of semantic knowledge derived from pretrained BERT when predicting the degrees of compositionality for 280 English noun compounds associated with human compositionality ratings. Our performance strongly improves on earlier unsupervised implementations of pretrained BERT and highlights beneficial decisions in data preprocessing, embedding computation, and compositionality estimation. The distinct linguistic roles of heads and modifiers are reflected by differences in BERT-derived representations, with empirical properties such as frequency, productivity, and ambiguity affecting model performance. The most relevant representational information is concentrated in the initial layers of the model architecture.

We present a novel dataset for physical and abstract plausibility of events in English. Based on naturally occurring sentences extracted from Wikipedia, we infiltrate degrees of abstractness, and automatically generate perturbed pseudo-implausible events. We annotate a filtered and balanced subset for plausibility using crowd-sourcing, and perform extensive cleansing to ensure annotation quality. In-depth quantitative analyses indicate that annotators favor plausibility over implausibility and disagree more on implausible events. Furthermore, our plausibility dataset is the first to capture abstractness in events to the same extent as concreteness, and we find that event abstractness has an impact on plausibility ratings: more concrete event participants trigger a perception of implausibility.

We investigate the effect of sub-word tokenization on representations of German noun compounds: single orthographic words which are composed of two or more constituents but often tokenized into units that are not morphologically motivated or meaningful. Using variants of BERT models and tokenization strategies on domain-specific restricted diachronic data, we introduce a suite of evaluations relying on the masked language modelling task and compositionality prediction. We obtain the most consistent improvements by pre-splitting compounds into constituents.

Humans tend to strongly agree on ratings on a scale for extreme cases (e.g., a CAT is judged as very concrete), but judgements on mid-scale words exhibit more disagreement. Yet, collected rating norms are heavily exploited across disciplines. Our study focuses on concreteness ratings and (i) implements correlations and supervised classification to identify salient multi-modal characteristics of mid-scale words, and (ii) applies a hard clustering to identify patterns of systematic disagreement across raters. Our results suggest to either fine-tune or filter mid-scale target words before utilising them.

Concrete words refer to concepts that are strongly experienced through human senses (banana, chair, salt, etc.), whereas abstract concepts are less perceptually salient (idea, glory, justice, etc.). A clear definition of abstractness is crucial for the understanding of human cognitive processes and for the development of natural language applications such as figurative language detection. In this study, we investigate selectional preferences as a criterion to distinguish between concrete and abstract concepts and words: we hypothesise that abstract and concrete verbs and nouns differ regarding the semantic classes of their arguments. Our study uses a collection of 5,438 nouns and 1,275 verbs to exploit selectional preferences as a salient characteristic in classifying English abstract vs. concrete words, and in predicting their concreteness scores. We achieve an f1-score of 0.84 for nouns and 0.71 for verbs in classification, and Spearman’s ρ correlation of 0.86 for nouns and 0.59 for verbs.

We provide a novel dataset – DiaWUG – with judgements on diatopic lexical semantic variation for six Spanish variants in Europe and Latin America. In contrast to most previous meaning-based resources and studies on semantic diatopic variation, we collect annotations on semantic relatedness for Spanish target words in their contexts from both a semasiological perspective (i.e., exploring the meanings of a word given its form, thus including polysemy) and an onomasiological perspective (i.e., exploring identical meanings of words with different forms, thus including synonymy). In addition, our novel dataset exploits and extends the existing framework DURel for annotating word senses in context (Erk et al., 2013; Schlechtweg et al., 2018) and the framework-embedded Word Usage Graphs (WUGs) – which up to now have mainly be used for semasiological tasks and resources – in order to distinguish, visualize and interpret lexical semantic variation of contextualized words in Spanish from these two perspectives, i.e., semasiological and onomasiological language variation.

Research on metaphorical language has shown ties between abstractness and emotionality with regard to metaphoricity; prior work is however limited to the word and sentence levels, and up to date there is no empirical study establishing the extent to which this is also true on the discourse level. This paper explores which textual and perceptual features human annotators perceive as important for the metaphoricity of discourses and expressions, and addresses two research questions more specifically. First, is a metaphorically-perceived discourse more abstract and more emotional in comparison to a literally- perceived discourse? Second, is a metaphorical expression preceded by a more metaphorical/abstract/emotional context than a synonymous literal alternative? We used a dataset of 1,000 corpus-extracted discourses for which crowdsourced annotators (1) provided judgements on whether they perceived the discourses as more metaphorical or more literal, and (2) systematically listed lexical terms which triggered their decisions in (1). Our results indicate that metaphorical discourses are more emotional and to a certain extent more abstract than literal discourses. However, neither the metaphoricity nor the abstractness and emotionality of the preceding discourse seem to play a role in triggering the choice between synonymous metaphorical vs. literal expressions. Our dataset is available at https://www.ims.uni-stuttgart.de/data/discourse-met-lit.

Agenda-setting is a widely explored phenomenon in political science: powerful stakeholders (governments or their financial supporters) have control over the media and set their agenda: political and economical powers determine which news should be salient. This is a clear case of targeted manipulation to divert the public attention from serious issues affecting internal politics (such as economic downturns and scandals) by flooding the media with potentially distracting information. We investigate agenda-setting in the Russian social media landscape, exploring the relation between economic indicators and mentions of foreign geopolitical entities, as well as of Russia itself. Our contributions are at three levels: at the level of the domain of the investigation, our study is the first to substructure the Russian media landscape in state-controlled vs. independent outlets in the context of strategic distraction from negative economic trends; at the level of the scope of the investigation, we involve a large set of geopolitical entities (while previous work has focused on the U.S.); at the qualitative level, our analysis of posts on Ukraine, whose relationship with Russia is of high geopolitical relevance, provides further insights into the contrast between state-controlled and independent outlets.

A variety of distributional and multi-modal computational approaches has been suggested for modelling the degrees of compositionality across types of multiword expressions and languages. As the starting point of my talk, I will present standard variants of computational models that have been proven successful in predicting the compositionality of German and English noun compounds. The main part of the talk will then be concerned with investigating the general reliability of these standard models and discussing implications for gold-standard datasets: I will demonstrate how prediction results vary (i) across representations, (ii) across empirical target properties, (iii) across compound types, (iv) across levels of abstractness, and (v) for general- vs. domain-specific language. Finally, I will present a preliminary quantitative study on diachronic changes of noun compound meanings and compositionality over time.

Given a specific discourse, which discourse properties trigger the use of metaphorical language, rather than using literal alternatives? For example, what drives people to say grasp the meaning rather than understand the meaning within a specific context? Many NLP approaches to metaphorical language rely on cognitive and (psycho-)linguistic insights and have successfully defined models of discourse coherence, abstractness and affect. In this work, we build five simple models relying on established cognitive and linguistic properties ? frequency, abstractness, affect, discourse coherence and contextualized word representations ? to predict the use of a metaphorical vs. synonymous literal expression in context. By comparing the models? outputs to human judgments, our study indicates that our selected properties are not sufficient to systematically explain metaphorical vs. literal language choices.

Type- and token-based embedding architectures are still competing in lexical semantic change detection. The recent success of type-based models in SemEval-2020 Task 1 has raised the question why the success of token-based models on a variety of other NLP tasks does not translate to our field. We investigate the influence of a range of variables on clusterings of BERT vectors and show that its low performance is largely due to orthographic information on the target word, which is encoded even in the higher layers of BERT representations. By reducing the influence of orthography we considerably improve BERT’s performance.

Kiezdeutsch is a variety of German predominantly spoken by teenagers from multi-ethnic urban neighborhoods in casual conversations with their peers. In recent years, the popularity of Kiezdeutsch has increased among young people, independently of their socio-economic origin, and has spread in social media, too. While previous studies have extensively investigated this language variety from a linguistic and qualitative perspective, not much has been done from a quantitative point of view. We perform the first large-scale data-driven analysis of the lexical and morpho-syntactic properties of Kiezdeutsch in comparison with standard German. At the level of results, we confirm predictions of previous qualitative analyses and integrate them with further observations on specific linguistic phenomena such as slang and self-centered speaker attitude. At the methodological level, we provide logistic regression as a framework to perform bottom-up feature selection in order to quantify differences across language varieties.

While there is a large amount of research in the field of Lexical Semantic Change Detection, only few approaches go beyond a standard benchmark evaluation of existing models. In this paper, we propose a shift of focus from change detection to change discovery, i.e., discovering novel word senses over time from the full corpus vocabulary. By heavily fine-tuning a type-based and a token-based approach on recently published German data, we demonstrate that both models can successfully be applied to discover new words undergoing meaning change. Furthermore, we provide an almost fully automated framework for both evaluation and discovery.

We suggest to model human-annotated Word Usage Graphs capturing fine-grained semantic proximity distinctions between word uses with a Bayesian formulation of the Weighted Stochastic Block Model, a generative model for random graphs popular in biology, physics and social sciences. By providing a probabilistic model of graded word meaning we aim to approach the slippery and yet widely used notion of word sense in a novel way. The proposed framework enables us to rigorously compare models of word senses with respect to their fit to the data. We perform extensive experiments and select the empirically most adequate model.

Predicting the difficulty of domain-specific vocabulary is an important task towards a better understanding of a domain, and to enhance the communication between lay people and experts. We investigate German closed noun compounds and focus on the interaction of compound-based lexical features (such as frequency and productivity) and terminology-based features (contrasting domain-specific and general language) across word representations and classifiers. Our prediction experiments complement insights from classification using (a) manually designed features to characterise termhood and compound formation and (b) compound and constituent word embeddings. We find that for a broad binary distinction into ‘easy’ vs. ‘difficult’ general-language compound frequency is sufficient, but for a more fine-grained four-class distinction it is crucial to include contrastive termhood features and compound and constituent features.

Predicting the degree of compositionality of noun compounds is a crucial ingredient for lexicography and NLP applications, to know whether the compound should be treated as a whole, or through its constituents. Computational approaches for an automatic prediction typically represent compounds and their constituents within a vector space to have a numeric relatedness measure for the words. This paper provides a systematic evaluation of using different vector-space reduction variants for the prediction. We demonstrate that Word2vec and nouns-only dimensionality reductions are the most successful and stable vector space reduction variants for our task.

We present a dataset with difficulty ratings for 1,030 German closed noun compounds extracted from domain-specific texts for do-it-ourself (DIY), cooking and automotive. The dataset includes two-part compounds for cooking and DIY, and two- to four-part compounds for automotive. The compounds were identified in text using the Simple Compound Splitter (Weller-Di Marco, 2017); a subset was filtered and balanced for frequency and productivity criteria as basis for manual annotation and fine-grained interpretation. This study presents the creation, the final dataset with ratings from 20 annotators and statistics over the dataset, to provide insight into the perception of domain-specific term difficulty. It is particularly striking that annotators agree on a coarse, binary distinction between easy vs. difficult domain-specific compounds but that a more fine grained distinction of difficulty is not meaningful. We finally discuss the challenges of an annotation for difficulty, which includes both the task description as well as the selection of the data basis.

Predicting the degree of compositionality of noun compounds such as “snowball” and “butterfly” is a crucial ingredient for lexicography and Natural Language Processing applications, to know whether the compound should be treated as a whole, or through its constituents, and what it means. Computational approaches for an automatic prediction typically represent and compare compounds and their constituents within a vector space and use distributional similarity as a proxy to predict the semantic relatedness between the compounds and their constituents as the compound’s degree of compositionality. This paper provides a systematic evaluation of vector-space reduction variants across kinds, exploring reductions based on part-of-speech next to and also in combination with Principal Components Analysis using Singular Value and word2vec embeddings. We show that word2vec and nouns only dimensionality reductions are the most successful and stable vector space variants for our task.

We perform a comparative study for automatic term extraction from domain-specific language using a PageRank model with different edge-weighting methods. We vary vector space representations within the PageRank graph algorithm, and we go beyond standard co-occurrence and investigate the influence of measures of association strength and first- vs. second-order co-occurrence. In addition, we incorporate meaning shifts from general to domain-specific language as personalized vectors, in order to distinguish between termhood strengths of ambiguous words across word senses. Our study is performed for two domain-specific English corpora: ACL and do-it-yourself (DIY); and a domain-specific German corpus: cooking. The models are assessed by applying average precision and the roc score as evaluation metrices.

Modelling language change is an increasingly important area of interest within the fields of sociolinguistics and historical linguistics. In recent years, there has been a growing number of publications whose main concern is studying changes that have occurred within the past centuries. The Corpus of Historical American English (COHA) is one of the most commonly used large corpora in diachronic studies in English. This paper describes methods applied to the downloadable version of the COHA corpus in order to overcome its main limitations, such as inconsistent lemmas and malformed tokens, without compromising its qualitative and distributional properties. The resulting corpus CCOHA contains a larger number of cleaned word tokens which can offer better insights into language change and allow for a larger variety of tasks to be performed.

While automatic term extraction is a well-researched area, computational approaches to distinguish between degrees of technicality are still understudied. We semi-automatically create a German gold standard of technicality across four domains, and illustrate the impact of a web-crawled general-language corpus on technicality prediction. When defining a classification approach that combines general-language and domain-specific word embeddings, we go beyond previous work and align vector spaces to gain comparative embeddings. We suggest two novel models to exploit general- vs. domain-specific comparisons: a simple neural network model with pre-computed comparative-embedding information as input, and a multi-channel model computing the comparison internally. Both models outperform previous approaches, with the multi-channel model performing best.

We present the results of our system for SemEval-2020 Task 1 that exploits a commonly used lexical semantic change detection model based on Skip-Gram with Negative Sampling. Our system focuses on Vector Initialization (VI) alignment, compares VI to the currently top-ranking models for Subtask 2 and demonstrates that these can be outperformed if we optimize VI dimensionality. We demonstrate that differences in performance can largely be attributed to model-specific sources of noise, and we reveal a strong relationship between dimensionality and frequency-induced noise in VI alignment. Our results suggest that lexical semantic change models integrating vector space alignment should pay more attention to the role of the dimensionality parameter.

We introduce SURel, a novel dataset with human-annotated meaning shifts between general-language and domain-specific contexts. We show that meaning shifts of term candidates cause errors in term extraction, and demonstrate that the SURel annotation reflects these errors. Furthermore, we illustrate that SURel enables us to assess optimisations of term extraction techniques when incorporating meaning shifts.

We perform an interdisciplinary large-scale evaluation for detecting lexical semantic divergences in a diachronic and in a synchronic task: semantic sense changes across time, and semantic sense changes across domains. Our work addresses the superficialness and lack of comparison in assessing models of diachronic lexical change, by bringing together and extending benchmark models on a common state-of-the-art evaluation task. In addition, we demonstrate that the same evaluation task and modelling approaches can successfully be utilised for the synchronic detection of domain-specific sense divergences in the field of term extraction.

Information about individuals can help to better understand what they say, particularly in social media where texts are short. Current approaches to modelling social media users pay attention to their social connections, but exploit this information in a static way, treating all connections uniformly. This ignores the fact, well known in sociolinguistics, that an individual may be part of several communities which are not equally relevant in all communicative situations. We present a model based on Graph Attention Networks that captures this observation. It dynamically explores the social graph of a user, computes a user representation given the most relevant connections for a target task, and combines it with linguistic information to make a prediction. We apply our model to three different tasks, evaluate it against alternative models, and analyse the results extensively, showing that it significantly outperforms other current methods.

In recent years, both cognitive and computational research has provided empirical analyses of contextual co-occurrence of concrete and abstract words, partially resulting in inconsistent pictures. In this work we provide a more fine-grained description of the distributional nature in the corpus-based interaction of verbs and nouns within subcategorisation, by investigating the concreteness of verbs and nouns that are in a specific syntactic relationship with each other, i.e., subject, direct object, and prepositional object. Overall, our experiments show consistent patterns in the distributional representation of subcategorising and subcategorised concrete and abstract words. At the same time, the studies reveal empirical evidence why contextual abstractness represents a valuable indicator for automatic non-literal language identification.

We simulate first- and second-order context overlap and show that Skip-Gram with Negative Sampling is similar to Singular Value Decomposition in capturing second-order co-occurrence information, while Pointwise Mutual Information is agnostic to it. We support the results with an empirical study finding that the models react differently when provided with additional second-order information. Our findings reveal a basic property of Skip-Gram with Negative Sampling and point towards an explanation of its success on a variety of tasks.

Cases of coreference and bridging resolution often require knowledge about semantic relations between anaphors and antecedents. We suggest state-of-the-art neural-network classifiers trained on relation benchmarks to predict and integrate likelihoods for relations. Two experiments with representations differing in noise and complexity improve our bridging but not our coreference resolver.

Automatic term identification and investigating the understandability of terms in a specialized domain are often treated as two separate lines of research. We propose a combined approach for this matter, by defining fine-grained classes of termhood and framing a classification task. The classes reflect tiers of a term’s association to a domain. The new setup is applied to German closed compounds as term candidates in the domain of cooking. For the prediction of the classes, we compare several neural network architectures and also take salient information about the compounds’ components into account. We show that applying a similar class distinction to the compounds’ components and propagating this information within the network improves the compound class prediction results.

Domain adaptation for sentiment analysis is challenging due to the fact that supervised classifiers are very sensitive to changes in domain. The two most prominent approaches to this problem are structural correspondence learning and autoencoders. However, they either require long training times or suffer greatly on highly divergent domains. Inspired by recent advances in cross-lingual sentiment analysis, we provide a novel perspective and cast the domain adaptation problem as an embedding projection task. Our model takes as input two mono-domain embedding spaces and learns to project them to a bi-domain space, which is jointly optimized to (1) project across domains and to (2) predict sentiment. We perform domain adaptation experiments on 20 source-target domain pairs for sentiment classification and report novel state-of-the-art results on 11 domain pairs, including the Amazon domain adaptation datasets and SemEval 2013 and 2016 datasets. Our analysis shows that our model performs comparably to state-of-the-art approaches on domains that are similar, while performing significantly better on highly divergent domains. Our code is available at https://github.com/jbarnesspain/domain_blse

Sentiment analysis in low-resource languages suffers from a lack of annotated corpora to estimate high-performing models. Machine translation and bilingual word embeddings provide some relief through cross-lingual sentiment approaches. However, they either require large amounts of parallel data or do not sufficiently capture sentiment information. We introduce Bilingual Sentiment Embeddings (BLSE), which jointly represent sentiment information in a source and target language. This model only requires a small bilingual lexicon, a source-language corpus annotated for sentiment, and monolingual word embeddings for each language. We perform experiments on three language combinations (Spanish, Catalan, Basque) for sentence-level cross-lingual sentiment classification and find that our model significantly outperforms state-of-the-art methods on four out of six experimental setups, as well as capturing complementary information to machine translation. Our analysis of the resulting embedding space provides evidence that it represents sentiment information in the resource-poor target language without any annotated data in that language.

We present a computational model to detect and distinguish analogies in meaning shifts between German base and complex verbs. In contrast to corpus-based studies, a novel dataset demonstrates that “regular” shifts represent the smallest class. Classification experiments relying on a standard similarity model successfully distinguish between four types of shifts, with verb classes boosting the performance, and affective features for abstractness, emotion and sentiment representing the most salient indicators.

We propose a framework that extends synchronic polysemy annotation to diachronic changes in lexical meaning, to counteract the lack of resources for evaluating computational models of lexical semantic change. Our framework exploits an intuitive notion of semantic relatedness, and distinguishes between innovative and reductive meaning changes with high inter-annotator agreement. The resulting test set for German comprises ratings from five annotators for the relatedness of 1,320 use pairs across 22 target words.

We present two novel datasets for the low-resource language Vietnamese to assess models of semantic similarity: ViCon comprises pairs of synonyms and antonyms across word classes, thus offering data to distinguish between similarity and dissimilarity. ViSim-400 provides degrees of similarity across five semantic relations, as rated by human judges. The two datasets are verified through standard co-occurrence and neural network models, showing results comparable to the respective English datasets.

This paper introduces a new dataset of term annotation. Given that even experts vary significantly in their understanding of termhood, and that term identification is mostly performed as a binary task, we offer a novel perspective to explore the common, natural understanding of what constitutes a term: Laypeople annotate single-word and multi-word terms, across four domains and across four task definitions. Analyses based on inter-annotator agreement offer insights into differences in term specificity, term granularity and subtermhood.

This paper presents two novel datasets and a random-forest classifier to automatically predict literal vs. non-literal language usage for a highly frequent type of multi-word expression in a low-resource language, i.e., Estonian. We demonstrate the value of language-specific indicators induced from theoretical linguistic research, which outperform a high majority baseline when combined with language-independent features of non-literal language (such as abstractness).

This paper presents a collection to assess meaning components in German complex verbs, which frequently undergo meaning shifts. We use a novel strategy to obtain source and target domain characterisations via sentence generation rather than sentence annotation. A selection of arrows adds spatial directional information to the generated contexts. We provide a broad qualitative description of the dataset, and a series of standard classification experiments verifies the quantitative reliability of the presented resource. The setup for collecting the meaning components is applicable also to other languages, regarding complex verbs as well as other language-specific targets that involve meaning shifts.

Across disciplines, researchers are eager to gain insight into empirical features of abstract vs. concrete concepts. In this work, we provide a detailed characterisation of the distributional nature of abstract and concrete words across 16,620 English nouns, verbs and adjectives. Specifically, we investigate the following questions: (1) What is the distribution of concreteness in the contexts of concrete and abstract target words? (2) What are the differences between concrete and abstract words in terms of contextual semantic diversity? (3) How does the entropy of concrete and abstract word contexts differ? Overall, our studies show consistent differences in the distributional representation of concrete and abstract words, thus challenging existing theories of cognition and providing a more fine-grained description of their nature.

Ambiguity represents an obstacle for distributional semantic models(DSMs), which typically subsume the contexts of all word senses within one vector. While individual vector space approaches have been concerned with sense discrimination (e.g., Schütze 1998, Erk 2009, Erk and Pado 2010), such discrimination has rarely been integrated into DSMs across semantic tasks. This paper presents a soft-clustering approach to sense discrimination that filters sense-irrelevant features when predicting the degrees of compositionality for German noun-noun compounds and German particle verbs.

This paper compares a neural network DSM relying on textual co-occurrences with a multi-modal model integrating visual information. We focus on nominal vs. verbal compounds, and zoom into lexical, empirical and perceptual target properties to explore the contribution of the visual modality. Our experiments show that (i) visual features contribute differently for verbs than for nouns, and (ii) images complement textual information, if (a) the textual modality by itself is poor and appropriate image subsets are used, or (b) the textual modality by itself is rich and large (potentially noisy) images are added.

Abstract words refer to things that can not be seen, heard, felt, smelled, or tasted as opposed to concrete words. Among other applications, the degree of abstractness has been shown to be a useful information for metaphor detection. Our contribution to this topic are as follows: i) we compare supervised techniques to learn and extend abstractness ratings for huge vocabularies ii) we learn and investigate norms for larger units by propagating abstractness to verb-noun pairs which lead to better metaphor detection iii) we overcome the limitation of learning a single rating per word and show that multi-sense abstractness ratings are potentially useful for metaphor detection. Finally, with this paper we publish automatically created abstractness norms for 3million English words and multi-words as well as automatically created sense specific abstractness ratings

There has been a good amount of progress in sentiment analysis over the past 10 years, including the proposal of new methods and the creation of benchmark datasets. In some papers, however, there is a tendency to compare models only on one or two datasets, either because of time restraints or because the model is tailored to a specific task. Accordingly, it is hard to understand how well a certain model generalizes across different tasks and datasets. In this paper, we contribute to this situation by comparing several models on six different benchmarks, which belong to different domains and additionally have different levels of granularity (binary, 3-class, 4-class and 5-class). We show that Bi-LSTMs perform well across datasets and that both LSTMs and Bi-LSTMs are particularly good at fine-grained sentiment tasks (i.e., with more than two classes). Incorporating sentiment information into word embeddings during training gives good results for datasets that are lexically similar to the training data. With our experiments, we contribute to a better understanding of the performance of different model architectures on different data sets. Consequently, we detect novel state-of-the-art results on the SenTube datasets.

We present a novel neural model HyperVec to learn hierarchical embeddings for hypernymy detection and directionality. While previous embeddings have shown limitations on prototypical hypernyms, HyperVec represents an unsupervised measure where embeddings are learned in a specific order and capture the hypernym–hyponym distributional hierarchy. Moreover, our model is able to generalize over unseen hypernymy pairs, when using only small sets of training data, and by mapping to other languages. Results on benchmark datasets show that HyperVec outperforms both state-of-the-art unsupervised measures and embedding models on hypernymy detection and directionality, and on predicting graded lexical entailment.

Distinguishing between antonyms and synonyms is a key task to achieve high performance in NLP systems. While they are notoriously difficult to distinguish by distributional co-occurrence models, pattern-based methods have proven effective to differentiate between the relations. In this paper, we present a novel neural network model AntSynNET that exploits lexico-syntactic patterns from syntactic parse trees. In addition to the lexical and syntactic information, we successfully integrate the distance between the related words along the syntactic path as a new pattern feature. The results from classification experiments show that AntSynNET improves the performance over prior pattern-based methods.

Up to date, the majority of computational models still determines the semantic relatedness between words (or larger linguistic units) on the type level. In this paper, we compare and extend multi-sense embeddings, in order to model and utilise word senses on the token level. We focus on the challenging class of complex verbs, and evaluate the model variants on various semantic tasks: semantic classification; predicting compositionality; and detecting non-literal language usage. While there is no overall best model, all models significantly outperform a word2vec single-sense skip baseline, thus demonstrating the need to distinguish between word senses in a distributional semantic model.

Many errors in phrase-based SMT can be attributed to problems on three linguistic levels: morphological complexity in the target language, structural differences and lexical choice. We explore combinations of linguistically motivated approaches to address these problems in English-to-German SMT and show that they are complementary to one another, but also that the popular verbal pre-ordering can cause problems on the morphological and lexical level. A discriminative classifier can overcome these problems, in particular when enriching standard lexical features with features geared towards verbal inflection.

Feature design and selection is a crucial aspect when treating terminology extraction as a machine learning classification problem. We designed feature classes which characterize different properties of terms based on distributions, and propose a new feature class for components of term candidates. By using random forests, we infer optimal features which are later used to build decision tree classifiers. We evaluate our method using the ACL RD-TEC dataset. We demonstrate the importance of the novel feature class for downgrading termhood which exploits properties of term components. Furthermore, our classification suggests that the identification of reliable term candidates should be performed successively, rather than just once.

This paper explores the information-theoretic measure entropy to detect metaphoric change, transferring ideas from hypernym detection to research on language change. We build the first diachronic test set for German as a standard for metaphoric change annotation. Our model is unsupervised, language-independent and generalizable to other processes of semantic change.

German particle verbs represent a frequent type of multi-word-expression that forms a highly productive paradigm in the lexicon. Similarly to other multi-word expressions, particle verbs exhibit various levels of compositionality. One of the major obstacles for the study of compositionality is the lack of representative gold standards of human ratings. In order to address this bottleneck, this paper presents such a gold standard data set containing 400 randomly selected German particle verbs. It is balanced across several particle types and three frequency bands, and accomplished by human ratings on the degree of semantic compositionality.

Vector space models and distributional information are widely used in NLP. The models typically rely on complex, high-dimensional objects. We present an interactive visualisation tool to explore salient lexical-semantic features of high-dimensional word objects and word similarities. Most visualisation tools provide only one low-dimensional map of the underlying data, so they are not capable of retaining the local and the global structure. We overcome this limitation by providing an additional trust-view to obtain a more realistic picture of the actual object distances. Additional tool options include the reference to a gold standard classification, the reference to a cluster analysis as well as listing the most salient (common) features for a selected subset of the words.

This paper presents a novel gold standard of German noun-noun compounds (Ghost-NN) including 868 compounds annotated with corpus frequencies of the compounds and their constituents, productivity and ambiguity of the constituents, semantic relations between the constituents, and compositionality ratings of compound-constituent pairs. Moreover, a subset of the compounds containing 180 compounds is balanced for the productivity of the modifiers (distinguishing low/mid/high productivity) and the ambiguity of the heads (distinguishing between heads with 1, 2 and >2 senses

This paper presents a collection of 350,000 German lemmatised words, rated on four psycholinguistic affective attributes. All ratings were obtained via a supervised learning algorithm that can automatically calculate a numerical rating of a word. We applied this algorithm to abstractness, arousal, imageability and valence. Comparison with human ratings reveals high correlation across all rating types. The full resource is publically available at: http://www.ims.uni-stuttgart.de/data/affective_norms/

Word embeddings have been demonstrated to benefit NLP tasks impressively. Yet, there is room for improvements in the vector representations, because current word embeddings typically contain unnecessary information, i.e., noise. We propose two novel models to improve word embeddings by unsupervised learning, in order to yield word denoising embeddings. The word denoising embeddings are obtained by strengthening salient information and weakening noise in the original word embeddings, based on a deep feed-forward neural network filter. Results from benchmark tasks show that the filtered word denoising embeddings outperform the original word embeddings.

We present a method for the extraction of synonyms for German particle verbs based on a word-aligned German-English parallel corpus: by translating the particle verb to a pivot, which is then translated back, a set of synonym candidates can be extracted and ranked according to the respective translation probabilities. In order to deal with separated particle verbs, we apply re-ordering rules to the German part of the data. In our evaluation against a gold standard, we compare different pre-processing strategies (lemmatized vs. inflected forms) and introduce language model scores of synonym candidates in the context of the input particle verb as well as distributional similarity as additional re-ranking criteria. Our evaluation shows that distributional similarity as a re-ranking feature is more robust than language model scores and leads to an improved ranking of the synonym candidates. In addition to evaluating against a gold standard, we also present a small-scale manual evaluation.

This paper addresses vector space models of prepositions, a notoriously ambiguous word class. We propose a rank-based distance measure to explore the vector-spatial properties of the ambiguous objects, focusing on two research tasks: (i) to distinguish polysemous from monosemous prepositions in vector space; and (ii) to determine salient vector-space features for a classification of preposition senses. The rank-based measure predicts the polysemy vs. monosemy of prepositions with a precision of up to 88%, and suggests preposition-subcategorised nouns as more salient preposition features than preposition-subcategorising verbs.

This paper discusses an extension of the V-measure (Rosenberg and Hirschberg, 2007), an entropy-based cluster evaluation metric. While the original work focused on evaluating hard clusterings, we introduce the Fuzzy V-measure which can be used on data that is inherently ambiguous. We perform multiple analyses varying the sizes and ambiguity rates and show that while entropy-based measures in general tend to suffer when ambiguity increases, a measure with desirable properties can be derived from these in a straightforward manner.

In the work presented here we assess the degree of compositionality of German Particle Verbs with a Distributional Semantics Model which only relies on word window information and has no access to syntactic information as such. Our method only takes the lexical distributional distance between the Particle Verb to its Base Verb as a predictor for compositionality. We show that the ranking of distributional similarity correlates significantly with the ranking of human judgements on semantic compositionality for a series of Particle Verbs and the Base Verbs they are derived from. We also investigate the influence of further linguistic factors, such as the ambiguity and the overall frequency of the verbs and a syntactically separate occurrences of verbs and particles that causes difficulties for the correct lemmatization of Particle Verbs. We analyse in how far these factors may influence the success with which the compositionality of the Particle Verbs may be predicted.

Translating prepositions is a difficult and under-studied problem in SMT. We present a novel method to improve the translation of prepositions by using noun classes to model their selectional preferences. We compare three variants of noun class information: (i) classes induced from the lexical resource GermaNet or obtained from clusterings based on either (ii) window information or (iii) syntactic features. Furthermore, we experiment with PP rule generalization. While we do not significantly improve over the baseline, our results demonstrate that (i) integrating selectional preferences as rigid class annotation in the parse tree is sub-optimal, and that (ii) clusterings based on window co-occurrence are more robust than syntax-based clusters or GermaNet classes for the task of modeling selectional preferences.

The current study works at the interface of theoretical and computational linguistics to explore the semantic properties of an particle verbs, i.e., German particle verbs with the particle an. Based on a thorough analysis of the particle verbs from a theoretical point of view, we identified empirical features and performed an automatic semantic classification. A focus of the study was on the mutual profit of theoretical and empirical perspectives with respect to salient semantic properties of the an particle verbs: (a) how can we transform the theoretical insights into empirical, corpus-based features, (b) to what extent can we replicate the theoretical classification by a machine learning approach, and (c) can the computational analysis in turn deepen our insights to the semantic properties of the particle verbs? The best classification result of 70% correct class assignments was reached through a GermaNet-based generalization of direct object nouns plus a prepositional phrase feature. These particle verb features in combination with a detailed analysis of the results at the same time confirmed and enlarged our knowledge about salient properties.

This paper introduces association norms of German noun compounds as a lexical semantic resource for cognitive and computational linguistics research on compositionality. Based on an existing database of German noun compounds, we collected human associations to the compounds and their constituents within a web experiment. The current study describes the collection process and a part-of-speech analysis of the association resource. In addition, we demonstrate that the associations provide insight into the semantic properties of the compounds, and perform a case study that predicts the degree of compositionality of the experiment compound nouns, as relying on the norms. Applying a comparatively simple measure of association overlap, we reach a Spearman rank correlation coefficient of rs=0.5228; p<000001, when comparing our predictions with human judgements.

There is by now widespread agreement that the most realistic way to construct the large-scale commonsense knowledge repositories required by natural language and artificial intelligence applications is by letting machines learn such knowledge from large quantities of data, like humans do. A lot of attention has consequently been paid to the development of increasingly sophisticated machine learning algorithms for knowledge extraction. However, the nature of the input that humans are exposed to while learning commonsense knowledge has received much less attention. The BabyExp project is collecting very dense audio and video recordings of the first 3 years of life of a baby. The corpus constructed in this way will be transcribed with automated techniques and made available to the research community. Moreover, techniques to extract commonsense conceptual knowledge incrementally from these multimodal data are also being explored within the project. The current paper describes BabyExp in general, and presents pilot studies on the feasibility of the automated audio and video transcriptions.

This paper presents a comparison of three computational approaches to selectional preferences: (i) an intuitive distributional approach that uses second-order co-occurrence of predicates and complement properties; (ii) an EM-based clustering approach that models the strengths of predicate--noun relationships by latent semantic clusters (Rooth et al., 1999); and (iii) an extension of the latent semantic clusters by incorporating the MDL principle into the EM training, thus explicitly modelling the predicate--noun selectional preferences by WordNet classes (Schulte im Walde et al., 2008). Concerning the distributional approach, we were interested not only in how well the model describes selectional preferences, but moreover which second-order properties are most salient. For example, a typical direct object of the verb 'drink' is usually fluid, might be hot or cold, can be bought, might be bottled, etc. The general question we ask is: what characterises the predicate's restrictions to the semantic realisation of its complements? Our second interest lies in the actual comparison of the models: How does a very simple distributional model compare to much more complex approaches, and which representation of selectional preferences is more appropriate, using (i) second-order properties, (ii) an implicit generalisation of nouns (by clusters), or (iii) an explicit generalisation of nouns by WordNet classes within clusters? We describe various experiments on German data and two evaluations, and demonstrate that the simple distributional model outperforms the more complex cluster-based models in most cases, but does itself not always beat the powerful frequency baseline.

Word sketches are part of the Sketch Engine corpus query system. They represent automatic, corpus-derived summaries of the words’ grammatical and collocational behaviour. Besides the corpus itself, word sketches require a sketch grammar, a regular expression-based shallow grammar over the part-of-speech tags, to extract evidence for the properties of the targeted words from the corpus. The paper presents a sketch grammar for German, a language which is not strictly configurational and which shows a considerable amount of case syncretism, and evaluates its accuracy, which has not been done for other sketch grammars. The evaluation focuses on NP case as a crucial part of the German grammar. We present various versions of NP definitions, so demonstrating the influence of grammar detail on precision and recall.

Distributional, corpus-based descriptions have frequently been applied to model aspects of word meaning. However, distributional models that use corpus data as their basis have one well-known disadvantage: even though the distributional features based on corpus co-occurrence were often successful in capturing meaning aspects of the words to be described, they generally fail to capture those meaning aspects that refer to world knowledge, because coherent texts tend not to provide redundant information that is presumably available knowledge. The question we ask in this paper is whether dictionary and encyclopaedic resources might complement the distributional information in corpus data, and provide world knowledge that is missing in corpora. As test case for meaning aspects, we rely on a collection of semantic associates to German verbs and nouns. Our results indicate that a combination of the knowledge resources should be helpful in work on distributional descriptions.

We describe a gold standard for semantic verb classes which is based on human associations to verbs. The associations were collected in a web experiment and then applied as verb features in a hierarchical cluster analysis. We claim that the resulting classes represent a theory-independent gold standard classification which covers a variety of semantic verb relations, and whose features can be used to guide the feature selection in automatic processes. To evaluate our claims, the association-based classification is validated against two standard approaches to semantic verb classes, GermaNet and FrameNet.