Subjects: Linguistics and Applied Linguistics >> Linguistics and Applied Linguistics Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2024-06-05
Abstract: Objective This paper aims to quantify the sentence alignment scores of low-resource parallel corpora to obtain high-quality parallel corpora, improving machine translation performance. Methods We propose NeuroAlign, a neural network-based unsupervised sentence embedding method for scoring bilingual parallel sentence alignment. Parallel sentence pairs are embedded into the same vector space, and alignment scores for given candidate sentence pairs in the parallel corpus are calculated. Based on these scores, low-scoring sentence pairs are filtered out, resulting in high-quality bilingual parallel corpora for low-resource languages. Results In the BUCC2018 parallel text mining task, the F1 score can be improved by 0.5-0.8. In the CCMT2021 low-resource language neural machine translation task, the BLEU score can be improved by 0.1-10.9. The sentence alignment scores can approach human evaluation. Limitations Due to the scarcity of low-resource bilingual parallel corpora, research has not been conducted on language pairs other than Tibetan-Chinese, Uyghur-Chinese, and Mongolian-Chinese. Conclusions This method can be effectively applied to sentence alignment scoring for low-resource language machine translation parallel corpora, improving the quality of the data source, and thereby enhancing machine translation performance.
Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2024-02-20
Abstract: OpenAI's Generative Pre-trained Transformer 4 (GPT-4) is a powerful large language model with a certain degree of intelligence in understanding and generating coherent text. We are exploring whether GPT-4 is capable of acting as a die, i.e. generating random numbers. We show that GPT-4 does not appear to generate independent and identically distributed random numbers. Examples imply that GPT-4 tries to compensate for the uniformity of random numbers by sacrificing independence when acting as a die.
Peer Review Status:Awaiting Review
Subjects: Linguistics and Applied Linguistics >> Linguistics and Applied Linguistics Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2024-01-11
Abstract: The research and engineering paradigm of natural language processing has been shifted with the rapid development of large languages models represented by the GPT series. It makes a significant impact on the related fields such as healthcare, education, judiciary and finance. At the same time, it also brings new possibilities for linguistics, the study of language itself. In this paper, we employ GPT4, Baichuan2 as well as ChatGLM3 and investigate their abilities of analyzing complex linguistic phenomena, taking ambiguity as an example. The experimental results show that GPT4 can effectively perceive and understand complex linguistic phenomena by integrating ambiguity resolution and syntactic analysis. For Baichuan2, if it is guided properly via prompt engineering, its analytical ability can be improved without parameter optimization. In addition, the relationship between linguistic phenomena and large language models can be visually demonstrated by monitoring the internal features and neuron activities of the models when processing ambiguous sentences in different context. In general, our experiments indicate that large language models are beneficial to better understanding the analyzing complex linguistic phenomena, hence providing new alternatives for linguistic research.
Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2024-01-07
Abstract: Multilayer perceptron (MLP) is a feedforward neural network that overcomes the limitations of linear models and opens the door to deep learning by adding one or more hidden layers to the network. In this paper, multilayer perceptrons are used to classfy image, which is explored on the Fashion MNIST dataset, and is attempted to be migrated to the MNIST dataset. In Fashion MNIST, we selected different optimization methods and compared them after feature preprocessing, optimized and improved the multi-layer perceptron by adding regularization methods such as dropout and weight decay.
Experiments show that appropriate feature processing can improve the numerical stability of the model. The momentum method significantly improves the effect of the model, weight decay and other regularization methods help to improve the generalization effect of the model.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2024-01-07
Abstract: The goals of this Chinese named entity recognition project mainly include the following two aspects. Firstly, it is to achieve high-precision Chinese named entity recognition. Through deep learning of Chinese text, the accuracy of Chinese entity recognition is improved, and the phenomenon of misidentification and missed recognition is reduced. Secondly, it is necessary to establish a standardized process and form a standardized Chinese named entity recognition process, including data preprocessing, model training, entity recognition, etc., to provide a foundation for subsequent research. The code has been submitted to GitHub at https://github.com/Blue88888/DL_CNER .
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2024-01-07
Abstract: During the literature review process, we focused on the work related to named entity recognition of the LSTM model, which has a wide range of applications in the field of named entity recognition. The English NER field has developed rapidly. As the project goal is to achieve Chinese named entity recognition, we discovered the gradual application of LSTM in the development of Chinese NER. LSTM is a variant of RNN, and its core concepts are cell state and "gate" structure.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2024-01-07
Abstract: In response to the current problems of inadequate and incomplete semantic feature extraction in Chinese named entity recognition research, Transformers (BERT) has shown striking improvements in a variety of related NLP tasks, and successive variants have been proposed to further improve the performance
of pre-trained language models. In this paper, our goal is to revisit Chinese pre-trained language models to examine their effectiveness in non-English languages. This paper is based on the RoBERT model for fine-tuning, and experimental results show good performance on many NLP tasks.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2024-01-05
Abstract: Based on the small sample problem, this article establishes conditional variational autoencoder network models for one-dimensional vibration signals and two-dimensional vibration signals, respectively. The sorted dataset is substituted into the model, and the good data generation ability of the generated model is utilized to generate new fault quantized signals of different types. Then, the generated component signals are reconstructed, and the data generation effect is preliminarily observed by comparing the original signal with the reconstructed signal, Then, the similarity between the reconstructed signal and the original signal is verified through dimensionality reduction visualization, cosine similarity, and maximum mean difference. Diversity is verified using parameters such as the maximum and minimum values of the signal, and the quality of the generated data is comprehensively evaluated. Finally, a fault information classification model was established, and the original sample sizes for three types of faults were set to 10, 30, and 50. The accuracy rates of fault information classification were 43.3%, 64.4%, and 84.7%, respectively. After sample expansion, the accuracy rate of fault information classification was 98.3%, fully verifying the accuracy of this method in generating fault data.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2024-01-04
Abstract: This paper explores potential directions for the application of large language models in the field ofYi studies, starting from the characteristics of Yi studies and the main scenarios for applying largelanguage models. It aims to investigate the potential role of large language models in Yi studies,highlighting specific applications of these models in Yi divination as an example. Yi studies havegarnered attention due to their extensive cultural significance, profound theoretical frameworks, andpractical applications, characterized by high specialization and a steep learning curve. Large languagemodels have achieved significant breakthroughs in natural language processing, knowledge graphconstruction, with major applications including machine translation, text summarization, intelligentquestion answering, and semantic understanding. However, the utilization of large language modelsin Yi studies remains largely underexplored. Therefore, this paper proposes potential directions forthe application of large language models in Yi studies, such as Yi literature analysis and interpretationof Yi divination results. Finally, by elaborating on a detailed application instance of large languagemodels in Yi divination, this paper reveals the potential of large language models in the field of Yistudies. A comprehensive understanding of the characteristics of Yi studies and extensive explorationof the application scenarios of large language models will contribute to fostering a broader applicationof these models in Yi studies, enhancing research efficiency, and promoting cultural inheritance.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2024-01-03
Abstract: League of Legends (LoL) is a highly popular multiplayer online competitive game, featuring intricate game mechanics and team cooperation, making the prediction of match outcomes a challenging task. This study utilizes a dataset from Kaggle, comprising 9,879 ranked matches ranging from Diamond I to Master tier, to build a machine learning model predicting the ultimate winner, either the blue or red team, based on the features of the first 10 minutes of gameplay. Through steps such as data loading, preprocessing, and feature engineering, we provided effective inputs for the model. For model selection, we opted for the Logistic Regression algorithm, achieving a model accuracy of 0.7277 through data splitting and training. This accuracy robustly supports predictions of the winning side, whether blue or red. However, to further enhance model performance, we recommend exploring additional feature en#2;gineering methods, investigating alternative machine learning algorithms, and fine-tuning hyperpa#2;rameters. The introduction of deep learning models is also a promising avenue to better capture the complex relationships within the game. Through these improvements, we anticipate increasing the model’s predictive accuracy for future matches, offering valuable insights for game development and enhancement.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2024-01-03
Abstract: Chinese Named Entity Recognition (NER) plays a key role in the field of natural language processing and is crucial for improving information extraction and text understanding. In this paper, we implement a Chinese named entity recognition method based on bi-directional long and short-term memory network (Bi-LSTM) and conditional random field (CRF). First, we extract features from the input Chinese text by Bi-LSTM network, and utilize the advantage of bidirectional sequence learning to capture contextual information for better understanding of the context. The network can effectively handle long-distance dependencies and improve the ability to recognize complex structures and nested entities in the text. Second, CRF is introduced as the decoding layer for sequence annotation task to globally optimize the output of Bi-LSTM to take into account the relationships between entity labels.The CRF model helps to correct local errors by capturing the global dependencies of labeled sequences, which improves the overall performance of the NER system. The experimental results show that the Bi-LSTM-CRF model performs well in the Chinese NER task and achieves good performance in all the metrics.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2023-12-17
Abstract: When training a large language model, the loss function will continue to decrease, making it difficult to determine the best stopping time. This article designs a fixed-length cross entropy so that the model loss will not decrease all the time. It will remain unchanged after the model is fully trained, which facilitates the selection of training stop time and saves training costs.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2023-11-01
Abstract: Most research and applications on natural language still concentrate on its superficial features and structures. However, natural language is essentially a way of encoding information and knowledge. Thus, the focus should be on what is encoded and how it is encoded. In line with this, we suggest a database-based approach for natural language processing that emulates the encoding of information and knowledge to build models. Based on these models, 1) generating sentences becomes akin to reading data from the models (or databases) and encoding it following some rules; 2) understanding sentences involves decoding rules and a series of boolean operations on the databases; 3) learning can be accomplished by writing on the databases. Our method closely mirrors how the human brain processes information, offering excellent interpretability and expandability.
Peer Review Status:Awaiting Review
Subjects: Management Science >> Enterprise Management Subjects: Computer Science >> Natural Language Understanding and Machine Translation Subjects: Library Science,Information Science >> Information Science Subjects: Other Disciplines >> Synthetic discipline submitted time 2023-10-31
Abstract: Purpose/significance ChatGPT is a chatbot program developed by OpenAI in the United States. A dialog with ChatGPT can provide insights into the theory and practice of think tanks. Method/process Currently, GPT-3.5 offers users a free query quota of 30 queries per day. Chen Yu has engaged in a dialog with ChatGPT on a number of issues related to the theory and practice of think tanks by creating an outline for the dialog. Result/conclusion AI technology, represented by ChatGPT, offers many opportunities for the think tank industry, including enhanced research capabilities, data-driven decision-making, and improved public engagement. However, it also poses challenges related to ethics, expertise, transparency, and workforce adaptability that think tanks need to seriously address. In the age of AI, Chinese think tanks and experts need to keep up with the trend and proactively adopt the AI technology represented by ChatGPT.
Peer Review Status:Awaiting Review
Subjects: Other Disciplines >> Synthetic discipline Subjects: Computer Science >> Natural Language Understanding and Machine Translation Subjects: Library Science,Information Science >> Information Science Subjects: Management Science >> Enterprise Management submitted time 2023-10-30
Abstract: Purpose/significance ChatGPT is a chatbot program developed by OpenAI in the United States. Conversations with ChatGPT can shed light on "Dialogue of Civilizations" in the age of AI. Method/process Currently, GPT-3.5 offers users 30 free query credits per day. By creating an outline for the conversation, Chen Yu engaged in a dialog with ChatGPT on various issues of "Dialogue of Civilizations". Result/conclusion Today, the "Standard of Civilization" has long been abandoned, and the "Clash of Civilizations" has been widely criticized. In the era of AI, the AI technology represented by ChatGPT can help promote the "Dialogue of Civilizations", help realize real-time communication between people of different cultural backgrounds, enhance the understanding and appreciation of different civilizations, and identify and alleviate prejudices in the dialogue of civilizations. At the same time, the AI technology represented by ChatGPT can also help promote "Dialogue within Civilizations" and play a positive role in resolving civil conflicts, promoting the integration of immigrants, protecting the voices of vulnerable groups, giving full play to the unique value of women, and building an age-friendly society. However, AI technologies must be developed and used with caution and with due regard to ethical considerations, in particular to prevent AI algorithms from perpetuating prejudices and reinforcing existing inequalities.
Peer Review Status:Awaiting Review
Subjects: Other Disciplines >> Synthetic discipline Subjects: Computer Science >> Natural Language Understanding and Machine Translation Subjects: Library Science,Information Science >> Information Science Subjects: Management Science >> Enterprise Management submitted time 2023-10-30
Abstract: Purpose/significance ChatGPT is a chatbot program developed by OpenAI in the United States. Conversations with ChatGPT can shed light on "Dialogue of Civilizations" in the age of AI. Method/process Currently, GPT-3.5 offers users 30 free query credits per day. By creating an outline for the conversation, Chen Yu engaged in a dialog with ChatGPT on various issues of "Dialogue of Civilizations". Result/conclusion Today, the "Standard of Civilization" has long been abandoned, and the "Clash of Civilizations" has been widely criticized. In the era of AI, the AI technology represented by ChatGPT can help promote the "Dialogue of Civilizations", help realize real-time communication between people of different cultural backgrounds, enhance the understanding and appreciation of different civilizations, and identify and alleviate prejudices in the dialogue of civilizations. At the same time, the AI technology represented by ChatGPT can also help promote "Dialogue within Civilizations" and play a positive role in resolving civil conflicts, promoting the integration of immigrants, protecting the voices of vulnerable groups, giving full play to the unique value of women, and building an age-friendly society. However, AI technologies must be developed and used with caution and with due regard to ethical considerations, in particular to prevent AI algorithms from perpetuating prejudices and reinforcing existing inequalities.
Peer Review Status:Awaiting Review
Subjects: Digital Publishing >> Digital News Subjects: Computer Science >> Natural Language Understanding and Machine Translation Subjects: Library Science,Information Science >> Information Science Subjects: Law >> Legal AI Subjects: Other Disciplines >> Synthetic discipline submitted time 2023-10-25
Abstract: Purpose/significance ChatGPT is a chatbot program developed by OpenAI in the United States. Conversations with ChatGPT can shed light on the media and communication industry in the age of AI. Method/process Currently, GPT-3.5 offers users 30 free query credits per day. By creating an outline for the conversation, Chen Yu engaged in a dialog with ChatGPT on various issues of the media and communication industry. Result/conclusion AI technology, represented by ChatGPT, has a huge impact on the media and communication industry. In the AI era, the media and communication industry should enthusiastically embrace AI technology and use it responsibly to provide a better experience for audiences. At the same time, the government, technology companies, civil society organizations, individuals, etc. should work together with the media and communication industry to solve the problems of fake news, cyber harassment, and information cocoon that may be brought about by AI technology.
Peer Review Status:Awaiting Review
Subjects: Digital Publishing >> Digital News Subjects: Computer Science >> Natural Language Understanding and Machine Translation Subjects: Library Science,Information Science >> Information Science Subjects: Law >> Legal AI Subjects: Other Disciplines >> Synthetic discipline submitted time 2023-10-25
Abstract: Purpose/significance ChatGPT is a chatbot program developed by OpenAI in the United States. Conversations with ChatGPT can shed light on the media and communication industry in the age of AI. Method/process Currently, GPT-3.5 offers users 30 free query credits per day. By creating an outline for the conversation, Chen Yu engaged in a dialog with ChatGPT on various issues of the media and communication industry. Result/conclusion AI technology, represented by ChatGPT, has a huge impact on the media and communication industry. In the AI era, the media and communication industry should enthusiastically embrace AI technology and use it responsibly to provide a better experience for audiences. At the same time, the government, technology companies, civil society organizations, individuals, etc. should work together with the media and communication industry to solve the problems of fake news, cyber harassment, and information cocoon that may be brought about by AI technology.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Natural Language Understanding and Machine Translation Subjects: Management Science >> Enterprise Management Subjects: Law >> Legal AI Subjects: Library Science,Information Science >> Information Science Subjects: Other Disciplines >> Synthetic discipline submitted time 2023-10-23
Abstract: Purpose/significance Today, countries around the world are accelerating their transformation to digital government. Conversations with ChatGPT can shed light on the digital government transformation in the age of AI. Method/process Currently, GPT-3.5 offers users 30 free query credits per day. By creating an outline for the conversation, Chen Yu engaged in a dialog with ChatGPT on various issues of the digital government transformation. Result/conclusion In the age of AI, AI technologies, such as ChatGPT, have the potential to revolutionize digital government transformation by increasing efficiency, improving service delivery, and enabling data-driven decision making. While the benefits are immense, governments must also address issues of ethics, bias, and workforce adaptation to ensure responsible and inclusive AI deployments that deliver better services and work outcomes for their citizens.
Peer Review Status:Awaiting Review
Subjects: Library Science,Information Science >> Information Science Subjects: Law >> Legal AI Subjects: Other Disciplines >> Synthetic discipline Subjects: Management Science >> Enterprise Management Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2023-10-23
Abstract: Purpose/significance Today, countries around the world are accelerating their transformation to digital government. Conversations with ChatGPT can shed light on the digital government transformation in the age of AI. Method/process Currently, GPT-3.5 offers users 30 free query credits per day. By creating an outline for the conversation, Chen Yu engaged in a dialog with ChatGPT on various issues of the digital government transformation. Result/conclusion In the age of AI, AI technologies, such as ChatGPT, have the potential to revolutionize digital government transformation by increasing efficiency, improving service delivery, and enabling data-driven decision making. While the benefits are immense, governments must also address issues of ethics, bias, and workforce adaptation to ensure responsible and inclusive AI deployments that deliver better services and work outcomes for their citizens.
Peer Review Status:Awaiting Review