In Korean, quantitative speech act studies have usually been conducted on single utterances with unspecified sources. In this study, we annotate sentences from the National Institute of Korean Language’s Messenger Corpus and the National Petition Corpus, as well as example sentences from an academic paper on contemporary Korean vlogging, and check the discrepancy between human annotation and model prediction. In particular, for sentences with differences in locutionary and illocutionary forces, we analyze the causes of errors to see if stylistic features used in a particular domain affect the correct inference of speech act. Through this, we see the necessity to build and analyze a balanced corpus in various text domains, taking into account cases with different usage roles, e.g., messenger conversations belonging to private conversations and petition corpus/vlogging script that have an unspecified audience.

Korean is a language with complex morphology that uses spaces at larger-than-word boundaries, unlike other East-Asian languages. While morpheme-based text generation can provide significant semantic advantages compared to commonly used character-level approaches, Korean morphological analyzers only provide a sequence of morpheme-level tokens, losing information in the tokenization process. Two crucial issues are the loss of spacing information and subcharacter level morpheme normalization, both of which make the tokenization result challenging to reconstruct the original input string, deterring the application to generative tasks. As this problem originates from the conventional scheme used when creating a POS tagging corpus, we propose an improvement to the existing scheme, which makes it friendlier to generative tasks. On top of that, we suggest a fully-automatic annotation of a corpus by leveraging public analyzers. We vote the surface and POS from the outcome and fill the sequence with the selected morphemes, yielding tokenization with a decent quality that incorporates space information. Our scheme is verified via an evaluation done on an external corpus, and subsequently, it is adapted to Korean Wikipedia to construct an open, permissive resource. We compare morphological analyzer performance trained on our corpus with existing methods, then perform an extrinsic evaluation on a downstream task.

Paraphrasing is often performed with less concern for controlled style conversion. Especially for questions and commands, style-variant paraphrasing can be crucial in tone and manner, which also matters with industrial applications such as dialog systems. In this paper, we attack this issue with a corpus construction scheme that simultaneously considers the core content and style of directives, namely intent and formality, for the Korean language. Utilizing manually generated natural language queries on six daily topics, we expand the corpus to formal and informal sentences by human rewriting and transferring. We verify the validity and industrial applicability of our approach by checking the adequate classification and inference performance that fit with conventional fine-tuning approaches, at the same time proposing a supervised formality transfer task.

Recently, with the advent of high-performance generative language models, artificial agents that communicate directly with the users have become more human-like. This development allows users to perform a diverse range of trials with the agents, and the responses are sometimes displayed online by users who share or show-off their experiences. In this study, we explore dialogues with a social chatbot uploaded to an online community, with the aim of understanding how users game human-like agents and display their conversations. Having done this, we assert that user postings can be investigated from two aspects, namely conversation topic and purpose of testing, and suggest a categorization scheme for the analysis. We analyze 639 dialogues to develop an annotation protocol for the evaluation, and measure the agreement to demonstrate the validity. We find that the dialogue content does not necessarily reflect the purpose of testing, and also that users come up with creative strategies to game the agent without being penalized.

There is an ongoing discussion on what makes humans more engaged when interacting with conversational agents. However, in the area of language processing, there has been a paucity of studies on how people react to agents and share interactions with others. We attack this issue by investigating the user dialogues with human-like agents posted online and aim to analyze the dialogue patterns. We construct a taxonomy to discern the users’ self-disclosure in the dialogue and the communication authenticity displayed in the user posting. We annotate the in-the-wild data, examine the reliability of the proposed scheme, and discuss how the categorization can be utilized for future research and industrial development.

In hate speech detection, developing training and evaluation datasets across various domains is the critical issue. Whereas, major approaches crawl social media texts and hire crowd-workers to annotate the data. Following this convention often restricts the scope of pejorative expressions to a single domain lacking generalization. Sometimes domain overlap between training corpus and evaluation set overestimate the prediction performance when pretraining language models on low-data language. To alleviate these problems in Korean, we propose APEACH that asks unspecified users to generate hate speech examples followed by minimal post-labeling. We find that APEACH can collect useful datasets that are less sensitive to the lexical overlaps between the pretraining corpus and the evaluation set, thereby properly measuring the model performance.

Social consensus has been established on the severity of online hate speech since it not only causes mental harm to the target, but also gives displeasure to the people who read it. For Korean, the definition and scope of hate speech have been discussed widely in researches, but such considerations were hardly extended to the construction of hate speech corpus. Therefore, we create a Korean online hate speech dataset with concrete annotation guideline to see how real world toxic expressions concern sociolinguistic discussions. This inductive observation reveals that hate speech in online news comments is mainly composed of social bias and toxicity. Furthermore, we check how the final corpus corresponds with the definition and scope of hate speech, and confirm that the overall procedure and outcome is in concurrence with the sociolinguistic discussions.

WARNING: This article contains contents that may offend the readers. Strategies that insert intentional noise into text when posting it are commonly observed in the online space, and sometimes they aim to let only certain community users understand the genuine semantics. In this paper, we explore the purpose of such actions by categorizing them into tricks, memes, fillers, and codes, and organize the linguistic strategies that are used for each purpose. Through this, we identify that such strategies can be conducted by authors for multiple purposes, regarding the presence of stakeholders such as ‘Peers’ and ‘Others’. We finally analyze how these strategies appear differently in each circumstance, along with the unified taxonomy accompanying examples.

In this technical report, we describe the fine-tuned ASR-MT pipeline used for the IWSLT shared task. We remove less useful speech samples by checking WER with an ASR model, and further train a wav2vec and Transformers-based ASR module based on the filtered data. In addition, we cleanse the errata that can interfere with the machine translation process and use it for Transformer-based MT module training. Finally, in the actual inference phase, we use a sentence boundary detection model trained with constrained data to properly merge fragment ASR outputs into full sentences. The merged sentences are post-processed using part of speech. The final result is yielded by the trained MT module. The performance using the dev set displays BLEU 20.37, and this model records the performance of BLEU 20.9 with the test set.

Assessing the similarity of sentences and detecting paraphrases is an essential task both in theory and practice, but achieving a reliable dataset requires high resource. In this paper, we propose a discourse component-based paraphrase generation for the directive utterances, which is efficient in terms of human-aided construction and content preservation. All discourse components are expressed in natural language phrases, and the phrases are created considering both speech act and topic so that the controlled construction of the sentence similarity dataset is available. Here, we investigate the validity of our scheme using the Korean language, a language with diverse paraphrasing due to frequent subject drop and scramblings. With 1,000 intent argument phrases and thus generated 10,000 utterances, we make up a sentence similarity dataset of practically sufficient size. It contains five sentence pair types, including paraphrase, and displays a total volume of about 550K. To emphasize the utility of the scheme and dataset, we measure the similarity matching performance via conventional natural language inference models, also suggesting the multi-lingual extensibility.

Korean is often referred to as a low-resource language in the research community. While this claim is partially true, it is also because the availability of resources is inadequately advertised and curated. This work curates and reviews a list of Korean corpora, first describing institution-level resource development, then further iterate through a list of current open datasets for different types of tasks. We then propose a direction on how open-source dataset construction and releases should be done for less-resourced languages to promote research.

Toxic comments in online platforms are an unavoidable social issue under the cloak of anonymity. Hate speech detection has been actively done for languages such as English, German, or Italian, where manually labeled corpus has been released. In this work, we first present 9.4K manually labeled entertainment news comments for identifying Korean toxic speech, collected from a widely used online news platform in Korea. The comments are annotated regarding social bias and hate speech since both aspects are correlated. The inter-annotator agreement Krippendorff’s alpha score is 0.492 and 0.496, respectively. We provide benchmarks using CharCNN, BiLSTM, and BERT, where BERT achieves the highest score on all tasks. The models generally display better performance on bias identification, since the hate speech detection is a more subjective issue. Additionally, when BERT is trained with bias label for hate speech detection, the prediction score increases, implying that bias and hate are intertwined. We make our dataset publicly available and open competitions with the corpus and benchmarks.

Modern dialog managers face the challenge of having to fulfill human-level conversational skills as part of common user expectations, including but not limited to discourse with no clear objective. Along with these requirements, agents are expected to extrapolate intent from the user’s dialogue even when subjected to non-canonical forms of speech. This depends on the agent’s comprehension of paraphrased forms of such utterances. Especially in low-resource languages, the lack of data is a bottleneck that prevents advancements of the comprehension performance for these types of agents. In this regard, here we demonstrate the necessity of extracting the intent argument of non-canonical directives in a natural language format, which may yield more accurate parsing, and suggest guidelines for building a parallel corpus for this purpose. Following the guidelines, we construct a Korean corpus of 50K instances of question/command-intent pairs, including the labels for classification of the utterance type. We also propose a method for mitigating class imbalance, demonstrating the potential applications of the corpus generation method and its multilingual extensibility.

Code-mixed grapheme-to-phoneme (G2P) conversion is a crucial issue for modern speech recognition and synthesis task, but has been seldom investigated in sentence-level in literature. In this study, we construct a system that performs precise and efficient multi-stage code-mixed G2P conversion, for a less studied agglutinative language, Korean. The proposed system undertakes a sentence-level transliteration that is effective in the accurate processing of Korean text. We formulate the underlying philosophy that supports our approach and demonstrate how it fits with the contemporary document.

Ethics regarding social bias has recently thrown striking issues in natural language processing. Especially for gender-related topics, the need for a system that reduces the model bias has grown in areas such as image captioning, content recommendation, and automated employment. However, detection and evaluation of gender bias in the machine translation systems are not yet thoroughly investigated, for the task being cross-lingual and challenging to define. In this paper, we propose a scheme for making up a test set that evaluates the gender bias in a machine translation system, with Korean, a language with gender-neutral pronouns. Three word/phrase sets are primarily constructed, each incorporating positive/negative expressions or occupations; all the terms are gender-independent or at least not biased to one side severely. Then, additional sentence lists are constructed concerning formality of the pronouns and politeness of the sentences. With the generated sentence set of size 4,236 in total, we evaluate gender bias in conventional machine translation systems utilizing the proposed measure, which is termed here as translation gender bias index (TGBI). The corpus and the code for evaluation is available on-line.

This paper proposes a novel feature extraction process for SemEval task 3: Irony detection in English tweets. The proposed system incorporates a concatenative featurization of tweet and hashtags, which helps distinguishing between the irony-related and the other components. The system embeds tweets into a vector sequence with widely used pretrained word vectors, partially using a character embedding for the words that are out of vocabulary. Identification was performed with BiLSTM and CNN classifiers, achieving F1 score of 0.5939 (23/42) and 0.3925 (10/28) each for the binary and the multi-class case, respectively. The reliability of the proposed scheme was verified by analyzing the Gold test data, which demonstrates how hashtags can be taken into account when identifying various types of irony.