Code-switching Data Research Articles

HC 2L: Hybrid and Cooperative Contrastive Learning for Cross-Lingual Spoken Language Understanding.

State-of-the-art model for zero-shot cross-lingual spoken language understanding performs cross-lingual unsupervised contrastive learning to achieve the label-agnostic semantic alignment between each utterance and its code-switched data. However, it ignores the precious intent/slot labels, whose label information is promising to help capture the label-aware semantics structure and then leverage supervised contrastive learning to improve both source and target languages' semantics. In this paper, we propose Hybrid and Cooperative Contrastive Learning to address this problem. Apart from cross-lingual unsupervised contrastive learning, we design a holistic approach that exploits source language supervised contrastive learning, cross-lingual supervised contrastive learning and multilingual supervised contrastive learning to perform label-aware semantics alignments in a comprehensive manner. Each kind of supervised contrastive learning mechanism includes both single-task and joint-task scenarios. In our model, one contrastive learning mechanism's input is enhanced by others. Thus the total four contrastive learning mechanisms are cooperative to learn more consistent and discriminative representations in the virtuous cycle during the training process. Experiments show that our model obtains consistent improvements over 9 languages, achieving new state-of-the-art performance.

Implikasi Penggunaan Alih Kode Dalam Film “Ngeri-Ngeri Sedap” Terhadap Komunikasi Antar Remaja di Lingkungan Sekolah

Language diversity allows each person to use more than one language. The use of several languages ​​in communication presents the potential for code mixing and code switching. These two linguistic events often occur in various environments, even in the mass media. This study aims to describe (1) the types and forms of code switching in the film Ngeri Ngeri Sedap; (2) factors that cause code switching in the film Ngeri-Ngeri Sedap; and (3) the use of the analysis results as a review of communication between teenagers in the school environment. This study is a qualitative descriptive study. The data collection technique used is the free listening technique with advanced techniques in the form of note-taking techniques. Data analysis is carried out through data reduction, data presentation, and drawing conclusions. The results of the study show that: (1) there are 15 internal code switching data involving Indonesian, Batak, Sundanese, and Javanese. (2) the results of the analysis can be used as a review of communication between teenagers in the school environment.

Language Options in Food Product Advertising on Youtube

The purpose of this study is to describe the form of language choice in speech events in advertisements for food products on YouTube and to describe the factors that determine the choice of language in speech events in advertisements for food products on YouTube.This research uses a qualitative descriptive research type. This research was carried out from December 2020 to May 2021. The data taken in this study were the utterances of commercial stars in 20 advertisements for food products on YouTube videos which were transcribed in written text. The data source used is food product advertisements on YouTube. The data collection technique used in this study is the note-taking technique. The data validation technique uses triangulation techniques. The method used in analyzing the data is the intralingual and extralingual equivalent methods. In food product advertisements on YouTube, 9 data were found using single language variations. Code switching data was found to be 2 codes. Then the ads using code mixing totaled 9 data with the form of words and phrases.

Deep Learning Approaches for English-Marathi Code-Switched Detection

During a conversation, speakers in multilingual societies frequently switch between two or more spoken languages. A linguistic action known as "code-switching" particularly alters or merges two or more languages. The development of software or tools for detecting code-switching has received very little attention. This paper proposes a Deep Learning based methods for detecting code-switched English-Marathi data. These suggested methods can be applied to various applications, including phone call merging, Intelligent AI assistants, Intelligent travelling systems to assist travellers in navigation and reservations, call centres to handle customer service issues, etc. To create a system for code switch detection, our study demonstrates a detailed analysis of extracting several audio features such as the Mel-Spectrogram, Mel-frequency Cepstral Coefficient (MFCC), and Perceptual Linear Predictive coefficients (PLP). Our team's English-Marathi code-switched dataset served as the testing ground for our methodologies. Our model's accuracy was 92.99%, with 40 MFCC coefficients having energy coefficient serving as the zeroth coefficient.

ALIH KODE DAN CAMPUR KODE PADA DIALOG ANTARTOKOH FILM BUMI MANUSIA KARYA HANUNG BRAMANTYO

This study aims to describe the forms of code-switching and code-mixing as well as the causal factors found in the film Earth of Mankind by Hanung Bramantyo. This research is a qualitative descriptive study using a sociolinguistic approach. The data collection technique used in this study is the observing and noting technique. The subject of this study is the film Bumi Manusia by Hanung Bramantyo. Based on the research that has been done, there are forms of code switching and code mixing as well as the causative factors that occur in the dialogue between the characters in the film Bumi Manusia by Hanung Bramantyo. The results of this study obtained 49 data, namely 11 code switching data and 38 code mixing data. In this case, there are several languages used, including Indonesian, English, Dutch, French and Javanese.

Code-switching input for machine translation: a case study of Vietnamese–English data

ABSTRACT Multilingualism presents both a challenge and an opportunity for Natural Language Processing, with code-switching representing a particularly interesting problem for computational models trained on monolingual datasets. In this paper, we explore how code-switched data affects the task of Machine Translation, a task which only recently has started to tackle the challenge of multilingual data. We test three Machine Translation systems on data from the Canberra Vietnamese–English Codeswitching Natural Speech Corpus (CanVEC) and evaluate translation output using both automatic and human metrics. We find that, perhaps counter-intuitively, Machine Translation performs better on code-switching input than monolingual input. In particular, comparison of human and automatic evaluation suggests that codeswitching input may boost the semantic faithfulness of the translation output, an effect we term lexico-semantic enrichment. We also report two cases where this effect is most and least clear in Vietnamese–English, namely gender-neutral 3SG pronouns and interrogative constructions respectively. Overall, we suggest that Machine Translation, and Natural Language Processing more generally, ought to view multilingualism as an opportunity rather than an obstacle.

An Analysis of Code-Switching on The Utterances of Agnes Mo on Daniel Mananta Network's Youtube Channel

This research aimed to analyze the phenomena of Code-Switching which examines the types and reasons for using code-switching used Agnes Monica on Daniel Mananta Network's YouTube channel entitled "Agnes Mo Pernah Punya Toxic Relationship?" which was published on February 11th, 2021. The researchers use Stockwell's (2002) theory that explains about three types of Code Switching, there are tag switching, intersetential switching, and intrasetential switching. In this research, the researchers employ a qualitative descriptive method, which is a research method that aims to explain and interpret reality and the researchers used Observation Method as the way to collect the data. The results and discussion found that there are six code-switching data, two Tag Switching data, two Intersetential Switching data, and two Intrasentential Switching data. And according to Hoffman (1991:116) in Saleh (2017:15), the researchers matched the data we found, that the reasons for the code-switching in Agnes Mo's utterances include the first one is talking about a particular topic, the second, being empathic about something, and the last the intention of clarifying the speech content for the interlocutor.

Types and Functions of Code-Switching in the Film Everything Everywhere All at Once

This case study analyzes the types and functions of code-switching (CS) used in a film entitled Everything Everywhere All at Once (EEAAO). Based on Poplack’s (1980) and Holmes’ (2012) researches, the researchers take a qualitative research method and a two-fold approach: counting the frequency of the different code switches with the analysis of whether there is a dominant type and interpreting the functions of CS examples. The result shows that all three types of CS, including inter-sentential, intra-sentential and extra-sentential switching, are used in EEAAO with no type that could be defined as a dominant type. As for functions, there are five of seven functions that appear in the CS data of EEAAO: solidarity, topic, switching for affective functions, lexical borrowing and metaphorical switching. The result indicates that CS in the film not only can be used as a node for emotional transformation and a thread for sorting out the plot but also as a means of language accommodation, embodying the integration of different cultures and enabling speakers to communicate more efficiently.

Alih Kode Pada Konten Vlog Dalam Kanal Youtube Turah Parthayana

The aims of the research are (1) to describe the forms of code switching contained in the content YouTube Turah Parthayana, (2) describes the form of code switching contained in the contentYouTube Turah Parthayana, (3) describes the factors that cause code switching in the contentYouTube To install Turah Parthayana.
 This research is a qualitative descriptive study. This research with sociolinguistic studies focuses on aspects of code switching. The data sources in this study are three video content shows owned by Turah Parthayana. The data collection technique in this study used the free-involved viewing method.
 Based on the results of the analysis, it can be concluded that the three video views are contentYouTube From that, the code switching data found was 44 data. The forms of temporary code switching and permanent code switching were both found by the researcher, namely 5 data. Forms of code switching found at the phrase level found 16 data, 10 data clause levels, 5 sentence levels, and 3 data inter-sentence levels. The factors that cause code-switching found are factors of talking about a certain topic, quoting other people, inserting interjections (expressive words of feeling, doing repetition used for clarification and intention to clarify the speaker's content to the other person

A Review and Critical Analysis of Qualitative Methodologies and Data-Collection Techniques Used for Code-Switching Research

<i>Code-switching is considered a language contact phenomenon that has long been a subject of scholarly investigation. This paper provides a review of common and nascent qualitative techniques often used to examine code-switching. Such a collection of salient qualitative methodologies could provide researchers with both prevailing and new frameworks and avenues for examining myriad types and aspects of code-switching. Numerous studies are highlighted with summaries of their data collection procedures, analysis techniques, and results. Also included are rationales for applying qualitative approaches to code-switching studies – especially poignant, as a trend to utilizing more experimental research designs have been recently observed regarding language contact phenomena. This paper will also discuss which types of research designs effectively combine techniques of acquiring code-switching data with rigorous approaches to data analysis to render new understandings of code-switching</i>

Kiswahili-English on Public Signage: A Morpheme- By -Morpheme Approach

Communication could be oral and written. These forms of communication may be formal or informal depending on the setting, people involved, age and class. Codeswitching is an informal form of communication. It is used most at times in oral communication. However, in recent times, its has been used extensively in written communication. This has raised issues as to whether written communication can also be a possible platform for informal communication. This study investigates codeswitching on public signage by Kiswahili –English bilinguals. The study examines how Kiswahili –English bilinguals use the two languages on public signage by structurally analysing how Kiswahili and English words are distributed in a given mixed constituent. To achieve this, Kiswahili-English codeswitched data was gathered from tokens of public signage from Kenya specifically, Nairobi. Using the Matrix Language Frame Model of Myers-Scotton, the study observes that theories propounded for oral codeswitching data could be used to analyse written communication (public signage texts) data because writers adhere to the syntax of both Kiswahili and English as stipulated by the tenets of the Matrix Language Frame Model. This is possible when the aim is to solely look at text items on public signage without keen considerations to the graphics therein that is, if the analysis is based purely on syntax. The study therefore recommends that the importance of codeswitching especially on public signage is evidence of how language is used and as accepted by members of a speech community.

The Analysis of the Sepedi-English Code-switched Radio News Corpus

Code-switching is a phenomenon that occurs mostly in multilingual countries where multilingual speakers often switch between languages in their conversations. The unavailability of large-scale code-switched corpora hampers the development and training of language models for the generation of code-switched text. In this study, we explore the initial phase of collecting and creating Sepedi-English code-switched corpus for generating synthetic news. Radio news and the frequency of code-switching on read news were considered and analysed. We developed and trained a Transformer-based language model using the collected code-switched dataset. We observed that the frequency of code-switched data in the dataset was very low at 1.1 %. We complemented our dataset with the news headlines dataset to create a new dataset. Although the frequency was still low, the model obtained the optimal loss rate of 2,361 with an accuracy of 66 %.

ANALYSIS OF CODE-SWITCHING AND CODE-MIXING USED IN RINTIK SEDU YOUTUBE CHANNEL’S VIDEO

<span lang="EN-GB">Rintik Sedu is an Indonesian writer's YouTube channel named Nadhifa Allya Tsana. As the content creator, Tsana masters Indonesian English and studies Korean self-taught. The mastery of the language causes code-mixing and code-switching found in Rintik Sedu YouTube channel’s videos. The study aimed to 1) identify the existing code-switching and code-mixing in Rintik Sedu YouTube channel’s video entitled </span><em><span lang="EN-GB">“it means you’re growing”</span></em><span lang="EN-GB">, 2) analyze the types of existing code-switching and code-mixing, and 3) describe the reason of code-switching and code-mixing used. The study used descriptive content analysis, which involved one of Rintik Sedu’s videos as a material to analyze. Some methods followed to collect the data, including searching the video, selecting and downloading the video, watching the video, transcribing the video’s audio, identifying the code-switching and code-mixing, and re-checking. The result found in this study are 1) Sixteen Indonesian Language-English code-switching data and fifteen Indonesian language-English code-mixing data. 2) The type of code-switching and code-mixing identified in Rintik Sedu’s video is external code-switching and outer code-mixing. 3) There are eight reasons for code-switching and code-mixing used.</span>

Heritage language maintenance and shift of three languages across three generations of Mountain Jews in Israel

ABSTRACT Mountain Jewish immigrants to Israel from the Eastern Caucasus used two heritage languages, Juhuri (Judeo-Tat) and Russian. Juhuri was their home and Russian the societal languages prior to migration. In Israel, Juhuri and Russian are Heritage Languages and Hebrew is the societal language. The present study reports on frequency of use and codeswitching behaviour across three generations of Mountain Jewish speakers in order to account for patterns of language maintenance and language shift (LMLS). Audio-recorded conversations were collected from six female middle-generation speakers (G1.5: ages 33–50) with their mothers (G1: ages 60–75) and children (G2: ages 9–21). Findings for language use showed significant cross-generational differences, where use of both Russian and Juhuri declined and use of Hebrew increased across generations. Juhuri was maintained only among G1 participants, G1.5 speakers using it mainly for comprehension and G2 speakers abandoning it entirely. G1.5 participants maintained Russian for interaction with parents but did not use it with their children, for whom Hebrew was the dominant means of communication. Code-switching data showed the same overall shift to Hebrew with some maintenance of Russian among G1.5 speakers. Results are discussed in terms of LMLS and motivations for code-switching.

Alih Kode Dan Campur Kode Dalam Video Blogger Suki Suka Japan

This research is entitled "Code-switching and code-mixing in the video blogger Suki Suka Japan". The purpose of this research is to represent the factors that cause and types of the process of code-switching and code-mixing on the video blogger Suki Suka Japan. This study uses three theories, namely the theory of code-switching (Wardhaugh); code-mixing theory (Hoffmann), and the theory that causes code-mixing (Weinrich). The research method is discriptive analysis with an approach using listening techniques and note-taking techniques. Data were analyzed using agih and tapping methods as well as informal. The results showed that there were 3 types of metaphorical code-switching data and two types of situational code-switching data. In the code-mixing data, there are 2 pieces (code-mixing in sentences), 3 pieces (mixing code in lexical), and 1 pieces (mixing code in phonological).

Building Educational Technologies for Code-Switching: Current Practices, Difficulties and Future Directions

Code-switching (CSW) is the phenomenon where speakers use two or more languages in a single discourse or utterance—an increasingly recognised natural product of multilingualism in many settings. In language teaching and learning in particular, code-switching has been shown to bring in many pedagogical benefits, including accelerating students’ confidence, increasing their access to content, as well as improving their participation and engagement. Unfortunately, however, current educational technologies are not yet able to keep up with this ‘multilingual turn’ in education, and are partly responsible for the constraint of this practice to only classroom contexts. In an effort to make progress in this area, we offer a data-driven position paper discussing the current state of affairs, difficulties of the existing educational natural language processing (NLP) tools for CSW and possible directions for future work. We specifically focus on two cases of feedback and assessment technologies, demonstrating how the current state-of-the-art in these domains fails with code-switching data due to a lack of appropriate training data, lack of robust evaluation benchmarks and lack of end-to-end user-facing educational applications. We present some empirical user cases of how CSW manifests and suggest possible technological solutions for each of these scenarios.

Adapter-based fine-tuning of pre-trained multilingual language models for code-mixed and code-switched text classification

Code-mixing and code-switching are frequent features in online conversations. Classification of such text is challenging if one of the languages is low-resourced. Fine-tuning pre-trained multilingual language models is a promising avenue for code-mixed text classification. In this paper, we explore adapter-based fine-tuning of PMLMs for CMCS text classification. We introduce sequential and parallel stacking of adapters, continuous fine-tuning of adapters, and training adapters without freezing the original model as novel techniques with respect to single-task CMCS text classification. We also present a newly annotated dataset for the classification of Sinhala–English code-mixed and code-switched text data, where Sinhala is a low-resourced language. Our dataset of 10000 user comments has been manually annotated for five classification tasks: sentiment analysis, humor detection, hate speech detection, language identification, and aspect identification, thus making it the first publicly available Sinhala–English CMCS dataset with the largest number of task annotation types. In addition to this dataset, we also tested our proposed techniques on Kannada–English and Hindi–English datasets. These experiments confirm that our adapter-based PMLM fine-tuning techniques outperform or are on par with the basic fine-tuning of PMLM models.

Code Mixing and Code Switching in Movie Murder on the Orient Express by Kenneth Branagh

This research aims to identify types of code-switching and code-mixing found in some of “Murder on The Orient Express” movie dialogues and utterances. The source of the data in this research is the dialogue in Murder on The Orient Express movie. To achieve the aims of this research; the researcher uses the descriptive qualitative research method. This study uses the code-switching theory of Wardhaugh (1986) and code-mixing theory of Suwito (1983). There are two types of code-switching consisting of metaphorical code-switching and situational code-switching and two types of code-mixing consisting of outer code-mixing and inner code-mixing. From the results of this study, the authors found 50 data of code-mixing and code-switching and as many as 4 or 8% metaphorical code-switching data, 2 or 4% situational code-switching data. The author also found 44 or 88% outer code-mixing data, and 0 or 0% inner code-mixing data. The research’s result shows that the dominant form of code-switching and code-mixing is metaphorical code-switching and outer code-mixing.

Bilinguals have a single computational system but two compartmentalized phonological grammars: Evidence from code-switching

Classic linguistic models, such as Chomsky’s minimalist schematization of the human language faculty, were typically based on a ‘monolingual ideal’. More recently, models have been extended to bilingual cognition. For instance, MacSwan (2000) posited that bilingual individuals possess a single syntactic computational system and, crucially, two phonological systems. The current paper examines this possible architecture of the bilingual language faculty by utilizing code-switching data. Specifically, the natural speech of Maria, a habitual Spanish-English code-switcher from the Bangor Miami Corpus, was examined. For the interface of phonology, an analysis was completed on the frequency of syllabic structures used by Maria. Phonotactics were examined as the (unilingual) phonological systems of Spanish and English impose differential restrictions on the legality of complex onsets and codas. The results indicated that Maria’s language of use impacted the phonotactics of her speech, but that the context of use (unilingual or code-switched) did not. This suggests that Maria was alternating between encapsulated phonological systems when she was code-switching. For the interface of morphosyntax, syntactic dependencies within Maria’s code-switched speech and past literature were examined. The evidence illustrates that syntactic dependencies are indeed established within code-switched sentences, indicating that such constructions are derived from a single syntactic subset. Thus, the quantitative and qualitative results from this paper wholly support MacSwan’s original conjectures regarding the bilingual language faculty: bilingual cognition appears to be composed of a single computational system which builds multi-language syntactic structures, and more than one phonological system.

COVID-19 and cyberbullying: deep ensemble model to identify cyberbullying from code-switched languages during the pandemic.

It has been declared by the World Health Organization (WHO) the novel coronavirus a global pandemic due to an exponential spread in COVID-19 in the past months reaching over 100 million cases and resulting in approximately 3 million deaths worldwide. Amid this pandemic, identification of cyberbullying has become a more evolving area of research over posts or comments in social media platforms. In multilingual societies like India, code-switched texts comprise the majority of the Internet. Identifying the online bullying of the code-switched user is bit challenging than monolingual cases. As a first step towards enabling the development of approaches for cyberbullying detection, we developed a new code-switched dataset, collected from Twitter utterances annotated with binary labels. To demonstrate the utility of the proposed dataset, we build different machine learning (Support Vector Machine & Logistic Regression) and deep learning (Multilayer Perceptron, Convolution Neural Network, BiLSTM, BERT) algorithms to detect cyberbullying of English-Hindi (En-Hi) code-switched text. Our proposed model integrates different hand-crafted features and is enriched by sequential and semantic patterns generated by different state-of-the-art deep neural network models. Initial experimental results of the proposed deep ensemble model on our code-switched data reveal that our approach yields state-of-the-art results, i.e., 0.93 in terms of macro-averaged F1 score. The dataset and codes of the present study will be made publicly available on the paper’s companion repository [https://github.com/95sayanta/COVID-19-and-Cyberbullying].