Hasil Pencarian  ::  Simpan CSV :: Kembali

"Penelitian ini mengusulkan pengembangan model monolingual untuk Natural Language Inference (NLI) dalam bahasa Swahili untuk mengatasi keterbatasan model multibahasa saat ini. Studi ini melakukan fine-tuning pada model SwahBERT yang sudah dilatih sebelumnya untuk menangkap hubungan semantik dan nuansa kontekstual unik dalam bahasa Swahili. Komponen penting dari penelitian ini adalah pembuatan dataset SwahiliNLI, yang dirancang untuk mencerminkan kompleksitas bahasa Swahili, sehingga menghindari ketergantungan pada teks bahasa Inggris yang diterjemahkan. Selain itu, kinerja model SwahBERT yang telah di-fine-tune dievaluasi menggunakan dataset SwahiliNLI dan XNLI, dan dibandingkan dengan model multibahasa mBERT. Hasilnya menunjukkan bahwa model SwahBERT mengungguli model multibahasa, mencapai tingkat akurasi sebesar 78,78% pada dataset SwahiliNLI dan 73,51% pada dataset XNLI. Model monolingual juga menunjukkan presisi, recall, dan skor F1 yang lebih baik, terutama dalam mengenali pola linguistik dan memprediksi pasangan kalimat. Penelitian ini menekankan pentingnya menggunakan dataset yang dihasilkan secara manual dan model monolingual dalam bahasa dengan sumber daya rendah, memberikan wawasan berharga untuk pengembangan sistem NLI yang lebih efisien dan relevan secara kontekstual, sehingga memajukan pemrosesan bahasa alami untuk bahasa Swahili dan berpotensi menguntungkan bahasa lain yang menghadapi keterbatasan sumber daya serupa.

---

This research proposes the development of a monolingual model for Natural Language Inference (NLI) in Swahili to overcome the limitations of current multilingual models. The study fine-tunes the pre-trained SwahBERT model to capture Swahili's unique semantic relationships and contextual nuances. A critical component of this research is the creation of a SwahiliNLI dataset, crafted to reflect the intricacies of the language, thereby avoiding reliance on translated English text. Furthermore, the performance of the fine-tuned SwahBERT model is evaluated using both SwahiliNLI and the XNLI dataset, and compared with the multilingual mBERT model. The results reveal that the SwahBERT model outperforms the multilingual model, achieving an accuracy rate of 78.78% on the SwahiliNLI dataset and 73.51% on the XNLI dataset. The monolingual model also exhibits superior precision, recall, and F1 scores, particularly in recognizing linguistic patterns and predicting sentence pairings. This research underscores the importance of using manually generated datasets and monolingual models in low-resource languages, providing valuable insights for the development of more efficient and contextually relevant NLI systems, thereby advancing natural language processing for Swahili and potentially benefiting other languages facing similar resource constraints."

"Margaret Masterman was a pioneer in the field of computational linguistics. Working in the earliest days of language processing by computer, she believed that meaning, not grammar, was the key to understanding languages, and that machines could determine the meaning of sentences. She was able, even on simple machines, to undertake sophisticated experiments in machine translation, and carried out important work on the use of semantic codings and thesauri to determine the meaning structure of text. This volume brings together Masterman’s groundbreaking papers for the first time. Through his insightful commentaries, Yorick Wilks argues that Masterman came close to developing a computational theory of language meaning based on the ideas of Wittgenstein, and shows the importance of her work in the philosophy of science and the nature of iconic languages. Of key interest in computational linguistics and artificial intelligence."

"Forum diskusi asinkron adalah salah satu media pembelajaran kolaboratif daring yang mampu mendorong pemikiran kritis, pertukaran gagasan, dan pembentukan pengetahuan. Analisis konten merupakan metode ilmiah yang dapat digunakan untuk mengidentifikasi keterampilan berpikir kritis dari transkrip pada forum diskusi asinkron. Metode analisis konten konvensional membutuhkan tahapan pengodean manual yang membutuhkan banyak waktu dan tenaga. Hal ini dapat mengakibatkan pengajar terlambat dalam memberikan intervensi instruksional karena informasi keterampilan berpikir kritis tidak dapat diperoleh secara cepat.Penelitian ini mengacu pada kerangka kerja Community of Inquiry (CoI) di mana keterampilan berpikir kritis dioperasionalisasikan melalui empat level dalam kehadiran kognitif yaitu pemantik diskusi, eksplorasi, integrasi, dan resolusi. Tujuan penelitian adalah mengembangkan model klasifikasi berbasis machine learning yang mampu menganalisis secara otomatis kehadiran kognitif pada transkrip diskusi berbahasa Indonesia. Desain penelitian menggunakan metode campuran kuantitatif dan kualitatif. Data eksperimen berjumlah 1.200 pesan diskusi dari mata kuliah Aljabar Linear di lingkungan pembelajaran bauran.Hasil penelitian menunjukkan bahwa kesiapan mahasiswa dalam mengelola pembelajaran dan lingkungan e-learning berpengaruh signifikan terhadap pengembangan kehadiran sosial dan kehadiran kognitif. Dataset level kehadiran kognitif pada transkrip diskusi asinkron dibangun dengan metode analisis konten yang reliabel kategori hampir sempurna (Cohen’s kappa = 0,88). Eksperimen pengembangan model analisis kehadiran kognitif menggunakan sepuluh basis algoritma yaitu XGBoost, Random Forest, Support Vector Machine, Logistic Regression, Naïve Bayes, Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), IndoBERT-base, IndoBERT-large dan XLM-RoBERTa. Model berbasis IndoBERT-large memiliki performa terbaik dengan akurasi sebesar 0,825. Prototipe sistem Cognipresa (cognitive presence analytics) telah dikembangkan untuk memfasilitasi pengajar dengan menganalisis kehadiran kognitif mahasiswa dalam diskusi secara otomatis. Evaluasi sistem menunjukkan hasil yang menjanjikan dari sisi usability dengan nilai System Usability Scale (SUS) sebesar 80,83.

---

The asynchronous discussion forum serves as a collaborative online learning platform capable of stimulating critical thinking, exchanging ideas, and shaping knowledge. Content analysis is a scientific method that can be employed to identify critical thinking skills from transcripts in asynchronous discussion forums. Conventional content analysis methods entail manual encoding stages, which consume a significant amount of time and effort. This may lead to instructors being delayed in providing instructional interventions due to the inability to swiftly obtain information on critical thinking skills.

This study references the Community of Inquiry (CoI) framework, where critical thinking skills are operationalized through four levels of cognitive presence: triggering event, exploration, integration, and resolution. The research's objective is to develop a machine learning-based classification model capable of automatically analyzing cognitive presence in Indonesian-language discussion transcripts. The research design incorporates both quantitative and qualitative methods. The experimental data consists of 1,200 discussion messages from the Linear Algebra course in a blended learning environment.

The research findings indicate that students' preparedness in managing learning and e-learning environment significantly influences the development of social presence and cognitive presence. The dataset for cognitive presence at the transcript of asynchronous discussions was constructed using a content analysis method with a reliably almost perfect category (Cohen’s kappa = 0.88). An experimental development of the cognitive presence analysis model was conducted using ten algorithmic bases, namely XGBoost, Random Forest, Support Vector Machine, Logistic Regression, Naïve Bayes, Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), IndoBERT-base, IndoBERT-large, and XLM- RoBERTa. The IndoBERT-large-based model demonstrated the best performance with an accuracy of 0.825. A prototype system called Cognipresa (cognitive presence analytics) has been developed to facilitate educators in automatically analyzing students' cognitive presence in discussions. The system evaluation indicates promising results in terms of usability, with a System Usability Scale (SUS) score of 80.83."

"This white paper is part of a series that promotes knowledge about language technology and its potential. This analysis focused on the 23 official European languages as well as other important national and regional languages in Europe. The results of this analysis suggest that there are many significant research gaps for each language. A more detailed expert analysis and assessment of the current situation will help maximise the impact of additional research and minimize any risks. META-NET consists of 54 research centres from 33 countries that are working with stakeholders from commercial businesses, government agencies, industry, research organisations, software companies, technology providers and European universities. Together, they are creating a common technology vision while developing a strategic research agenda that shows how language technology applications can address any research gaps by 2020."

"This white paper is part of a series that promotes knowledge about language technology and its potential. This analysis focused on the 23 official European languages as well as other important national and regional languages in Europe. The results of this analysis suggest that there are many significant research gaps for each language. A more detailed expert analysis and assessment of the current situation will help maximise the impact of additional research and minimize any risks. META-NET consists of 54 research centres from 33 countries that are working with stakeholders from commercial businesses, government agencies, industry, research organisations, software companies, technology providers and European universities. Together, they are creating a common technology vision while developing a strategic research agenda that shows how language technology applications can address any research gaps by 2020."

"This white paper is part of a series that promotes knowledge about language technology and its potential. This analysis focused on the 23 official European languages as well as other important national and regional languages in Europe. The results of this analysis suggest that there are many significant research gaps for each language. A more detailed expert analysis and assessment of the current situation will help maximise the impact of additional research and minimize any risks. META-NET consists of 54 research centres from 33 countries that are working with stakeholders from commercial businesses, government agencies, industry, research organisations, software companies, technology providers and European universities. Together, they are creating a common technology vision while developing a strategic research agenda that shows how language technology applications can address any research gaps by 2020."

"This white paper is part of a series that promotes knowledge about language technology and its potential. This analysis focused on the 23 official European languages as well as other important national and regional languages in Europe. The results of this analysis suggest that there are many significant research gaps for each language. A more detailed expert analysis and assessment of the current situation will help maximise the impact of additional research and minimize any risks. META-NET consists of 54 research centres from 33 countries that are working with stakeholders from commercial businesses, government agencies, industry, research organisations, software companies, technology providers and European universities. Together, they are creating a common technology vision while developing a strategic research agenda that shows how language technology applications can address any research gaps by 2020"

"This white paper is part of a series that promotes knowledge about language technology and its potential. This analysis focused on the 23 official European languages as well as other important national and regional languages in Europe. The results of this analysis suggest that there are many significant research gaps for each language. A more detailed expert analysis and assessment of the current situation will help maximise the impact of additional research and minimize any risks. META-NET consists of 54 research centres from 33 countries that are working with stakeholders from commercial businesses, government agencies, industry, research organisations, software companies, technology providers and European universities. Together, they are creating a common technology vision while developing a strategic research agenda that shows how language technology applications can address any research gaps by 2020. ​"