Natural Language Processing | 自然語言處理

  • What is it?
    Natural language processing, often abbreviated as NLP, concerns the training of artificial intelligence to comprehend human speech and text in much the same way that human beings do. To wit, humans are prone to mispronunciations, misspellings, and grammatical errors; what's more, what a word or sentence means depends largely on the context. For instance, humans can understand sarcasm and hyperbole without much issue, and intentional misspellings and broken grammar have become par for the course in the era of internet-based communication. For the AI, however, this used to constitute an insurmountable hurdle.

    With the advancement of machine learning and deep learning methods, along with a scientific and engineering discipline known as computational linguistics, modern AI applications have become much more apt at understanding humans. Many so-called "smart" devices and services, such as speech recognition and machine translation, rely on NLP to accurately interpret your meaning and carry out your instructions.

  • Why do you need it?
    Leaving aside the fact that most of us are already benefiting from NLP in our daily lives, the implementation of NLP in a business environment can be a game-changer. In marketing and customer service, spam detection, social listening, and chatbots can ensure your company really hears what your customers are saying, so you can engage with them in a meaningful manner. Other potential applications include automatic summarization of long swaths of text, natural language generation and understanding (NLG/NLU), and other more advanced AI applications.

  • How is GIGABYTE helpful?
    Since machine learning and deep learning figure prominently in the development of NLP, GIGABYTE's hardware and software solutions can be an important asset. For example, GIGABYTE's G-Series GPU Servers support CPUs with GPGPU accelerators to provide HPC capabilities that can be used to achieve deep learning and MLOps. The Myelintek MLSteam DNN Training System, which is part of GIGABYTE's DNN Training Appliance, provides a turnkey DNN training platform that can also be run on GIGABYTE's GPU Servers. Another way GIGABYTE is contributing to NLP is an interesting case of GIGABYTE's long-term CSR and ESG efforts being expressed through its advanced server solutions. In 2020, GIGABYTE provided four G482-Z50 servers for Taiwan's Cheng Kung University (NCKU), which was participating in the APAC HPC-AI Competition being held in Singapore. The NCKU team went on to win first place and break NLP world records using BERT, a machine learning technique developed by Google. The team achieved an accuracy of 87.7%, which was even higher than what had been achieved by the University of California, San Diego and Stanford University at the time.

    Learn more : 《Advance AI with GIGABYTE’s supercharged AI server solutions

  • WE RECOMMEND
    RELATED ARTICLES
    如何將人工智慧導入醫療保健業

    AI & AIoT

    如何將人工智慧導入醫療保健業

    從事醫療與健康產業的讀者,請花幾分鐘閱讀本篇文章,了解人工智慧(AI)為這個領域所帶來的嶄新商機,並認識能助您從中受益的AI工具。本篇是技嘉科技「Power of AI」系列文章,目的是介紹不同產業的最新AI趨勢,鼓勵前瞻者把握AI浪潮,找尋自己的機會。
    如何挑選您的AI伺服器?(上)CPU和GPU

    Tech Guide

    如何挑選您的AI伺服器?(上)CPU和GPU

    生成式人工智慧和其他AI工具盛行的當下,挑選合適的AI伺服器成為各產業的首要任務。技嘉科技最新發表《科技指南》,帶領讀者認識AI伺服器的八個關鍵零組件,本篇從最重要的元件開始,即中央處理器(CPU)和圖形處理器(GPU)。挑選適當的運算晶片,打造量身訂做的人工智慧超算平台,可以讓工作事半功倍,為使用者開創全新的巔峰。
    帶您快速跟上人工智慧AI趨勢的十大問答

    AI & AIoT

    帶您快速跟上人工智慧AI趨勢的十大問答

    大家都在談人工智慧(AI),您是否也希望擁有基本的知識,參與這個話題的討論?別擔心,技嘉科技為您準備了介紹AI趨勢的十大問答,讓您能快速理解人工智慧的概念!
    揭秘生成式AI:從「訓練」到「推論」,一窺其神奇運作原理

    Tech Guide

    揭秘生成式AI:從「訓練」到「推論」,一窺其神奇運作原理

    生成式人工智慧(generative AI)來勢洶洶,實用案例包括聊天機器人ChatGPT能撰寫文案與企劃案、Stable Diffusion等圖像生成工具能依指令產出栩栩如生的影像。事實上,想搞懂生成式AI不難,使用它比想像中還容易。技嘉科技在本篇《科技指南》中,為您解剖生成式AI背後「訓練」與「推論」兩個重要步驟,並為您推薦專為人工智慧開發所設計的技嘉解決方案,協助您成為AI世代的佼佼者。
    CPU vs. GPU:淺談伺服器的兩大運算力

    Tech Guide

    CPU vs. GPU:淺談伺服器的兩大運算力

    順應科技趨勢,業界持續追求運算力更強大的伺服器,其中的關鍵元素除了大家熟知的中央處理器CPU之外,圖形處理器GPU近年來也受到重視;但您是否知道,什麼是GPU?它與CPU的差別何在?又該如何運用兩種不同類別的處理器,發揮最強大的運算力?技嘉科技是伺服器與尖端科技解決方案的知名品牌,發表本篇《科技指南》,目的是比較CPU與GPU的原理與功能,進一步協助您挑選合適的技嘉伺服器產品,將「GPU運算力」導入您的伺服器機房與資料中心。