This report summarizes the objectives and evaluation of SemEval 2015 task on sentiment analysis figurative language Twitter (Task 11).This is first wholly dedicated to analyzing Twitter.Specifically, three broad classes are considered: irony, sarcasm metaphor.Gold standard sets 8000 training tweets 4000 test were annotated using workers crowdsourcing platform CrowdFlower.Participating systems required provide a fine-grained score an 11-point scale (-5 +5, including 0 for neutral intent) each...

Metaphor is pervasive in our communication, which makes it an important problem for natural language processing (NLP).Numerous approaches to metaphor have thus been proposed, all of relied on linguistic features and textual data construct their models.Human comprehension is, however, known rely both perceptual experience, vision can play a particularly role when metaphorically projecting imagery across domains.In this paper, we present the first identification method that simultaneously...

It is generally believed that a metaphor tends to have stronger emotional impact than literal statement; however, there no quantitative study establishing the extent which this true.Further, mechanisms through metaphors convey emotions are not well understood.We present first data-driven comparing emotionality of metaphorical expressions with their counterparts.Our results indicate usages are, on average, significantly more usages.We also show content simply transferred from source domain...

Linguistic typology aims to capture structural and semantic variation across the world’s languages. A large-scale could provide excellent guidance for multilingual Natural Language Processing (NLP), particularly languages that suffer from lack of human labeled resources. We present an extensive literature survey on use typological information in development NLP techniques. Our demonstrates date, existing databases has resulted consistent but modest improvements system performance. show this...

Metaphor is highly frequent in language, which makes its computational processing indispensable for real-world NLP applications addressing semantic tasks. Previous approaches to metaphor modeling rely on task-specific hand-coded knowledge and operate a limited domain or subset of phenomena. We present the first integrated open-domain statistical model unrestricted text. Our method identifies metaphorical expressions running text then paraphrases them with their literal paraphrases. Such...

As the community working on computational approaches to figurative language is growing and as methods data become increasingly diverse, it important create widely shared empirical knowledge of level system performance in a range contexts, thus facilitating progress this area. One way creating such through benchmarking multiple systems common dataset. We report task metaphor identification VU Amsterdam Metaphor Corpus conducted at NAACL 2018 Workshop Figurative Language Processing.

Beata Beigman Klebanov, Chee Wee Leong, E. Dario Gutierrez, Ekaterina Shutova, Michael Flor. Proceedings of the 54th Annual Meeting Association for Computational Linguistics (Volume 2: Short Papers). 2016.

System design and evaluation methodologies receive significant attention in natural language processing (NLP), with the systems typically being evaluated on a common task against shared data sets. This enables direct system comparison facilitates progress field. However, computational work metaphor is considerably more fragmented than similar research efforts other areas of NLP semantics. Recent years have seen growing interest modeling metaphor, many new statistical techniques opening...

The ubiquity of metaphor in our everyday communication makes it an important problem for natural language understanding. Yet, the majority processing systems to date rely on hand-engineered features and there is still no consensus field as which are optimal this task. In paper, we present first deep learning architecture designed capture metaphorical composition. Our results demonstrate that outperforms existing approaches identification

Metaphorical expressions are pervasive in natural language and pose a substantial challenge for computational semantics.The inherent compositionality of metaphor makes it an important test case compositional distributional semantic models (CDSMs).This paper is the first to investigate whether metaphorical composition warrants distinct treatment CDSM framework.We propose method learn metaphors as linear transformations vector space find that, across variety domains, explicitly modeling...

Verna Dankers, Marek Rei, Martha Lewis, Ekaterina Shutova. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint (EMNLP-IJCNLP). 2019.

Abuse on the Internet represents an important societal problem of our time. Millions users face harassment, racism, personal attacks, and other types abuse online platforms. The psychological effects such individuals can be profound lasting. Consequently, over past few years, there has been a substantial research effort towards automated detection in field natural language processing (NLP). In this paper, we present comprehensive survey methods that have proposed to date, thus providing...

The rise of online communication platforms has been accompanied by some undesirable effects, such as the proliferation aggressive and abusive behaviour online. Aiming to tackle this problem, natural language processing (NLP) community experimented with a range techniques for abuse detection. While achieving substantial success, these methods have so far only focused on modelling linguistic properties comments communities users, disregarding emotional state users how might affect their...

The advent of social media in recent years has fed into some highly undesirable phenomena such as proliferation offensive language, hate speech, sexist remarks, etc. on the Internet. In light this, there have been several efforts to automate detection and moderation abusive content. However, deliberate obfuscation words by users evade poses a serious challenge effectiveness these efforts. current state art approaches language detection, based recurrent neural networks, do not explicitly...

Highly frequent in language and communication, metaphor represents a significant challenge for Natural Language Processing (NLP) applications. Computational work on has traditionally evolved around the use of hand-coded knowledge, making systems hard to scale. Recent years have witnessed rise statistical approaches processing. However, these often require extensive human annotation effort are predominantly evaluated within limited domain. In contrast, we experiment with weakly supervised...

This book offers a comprehensive approach to the computational treatment of metaphor and its figurative brethren.

An increasingly common expression of online hate speech is multimodal in nature and comes the form memes. Designing systems to automatically detect hateful content paramount importance if we are mitigate its undesirable effects on society at large. The detection an intrinsically difficult open problem: memes convey a message using both images text and, hence, require reasoning joint visual language understanding. In this work, seek advance line research develop framework for We improve...