دانلود مقاله هوش وب و انفورماتیک مغز: آینده هوش مصنوعی
چکیده
مقدمه
هوش مصنوعی در دنیای متصل
روندها و چالش ها
نتایجی اظهار شده
منابع
Abstract
Introduction
AI in the connected world
Trends and challenges
Concluding remarks
References
چکیده
درک هوش انسانی، به ویژه هوش مغز، سنگ بنای رسیدن به هوش مصنوعی نهایی است. در این مقاله، به طور خلاصه تعاملات تاریخی بین هوش مصنوعی و علم مغز را مرور می‌کنیم و به چشم‌انداز آینده هوش مصنوعی در دنیای متصل نگاه می‌کنیم. به طور خاص، ما دو زمینه به سرعت در حال توسعه در هوش وب (WI، AI در جهان متصل) و انفورماتیک مغز (BI، مطالعه مغز/ذهن محوری و کاربرد هوش مغز-ماشین) را معرفی کرده و آنها را برای تسریع ورود جامعه هوش مصنوعی در سطح انسانی علاوه بر این، ترکیب این دو زمینه با اتصال هوش مصنوعی و علم مغز با داده‌های بزرگ، چشم‌انداز جدیدی از تحقیقات سیستماتیک هوش مغز-ماشین تا زنجیره صنعت جدید هوش مصنوعی در فضاهای مرتبط اجتماعی-سایبری-فیزیکی-فکری ایجاد می‌کند.
توجه! این متن ترجمه ماشینی بوده و توسط مترجمین ای ترجمه، ترجمه نشده است.
Abstract
Understanding human intelligence, especially brain intelligence, is the cornerstone of reaching the ultimate AI. In this paper, we briefly review the historical interactions between AI and brain science, and look towards the future vision of AI in the connected world. In particular, we introduce two rapidly developing fields in Web Intelligence (WI, AI in the Connected World) and Brain Informatics (BI, the brain/mind-centric study and the application of brain-machine intelligence), and combine them to accelerate the arrival of a human–level AI society. Furthermore, combining these two fields by connecting AI and brain science with big data, creates a new vision from the systematic brain-machine intelligence research to new AI industry chain in the connected social-cyber-physical-thinking spaces.
Introduction
From its conception, artifcial intelligence (AI) has experienced several key milestones, each of which had its own topics that inspired new development trends enriching scientifc and technological progress. More specifcally, the perception capability was focused in the early period. Subsequently, the inference-centric study led AI to developing the frstgeneration robotic and intelligent software. For instance, the structure learning and inductive learning systems were developed as the main viewpoint, at that time. In this period of knowledge, the expert system was studied as the mainstream. Meanwhile, neural networks also achieved a breakthrough, especially for multilayer perceptron and backpropagation. In the third period to now, machine learning and deep learning have become the mainstream, achieving remarkable gains in many felds such as pattern recognition, neural language processing and control [52, 83]. In this period, scientists focus on how to make machines own the higher-level learning capability, that is, human intelligence and the cognitive capability. In this context, many supercomputers, intelligent robots and online applications were developed in academia and industry, such as Deep Blue [76], the NAO humanoid robots [82] and Alpha Go[89]. When it comes to developing the intelligence techniques, brain decoding is widely regarded as its fundamental and essential roads. By understanding biological characteristics and brain information-processing mechanisms, the intelligence capability is developed, modeled, simulated and assigned into machines, empowering them to become more humane. It is a long history to develop the brain-inspired intelligence applications, especially from the bottom-up perspective. However, owing to the limit of decoding brain, its support for AI seems rather slow. In recent years, with the development of new technologies and methodologies, the brain intelligence study has become a hot topic again. We argue that this topic will keep leading trends of AI study. For this, the core issue is how to narrow the gap between brain science and artifcial intelligence towards achieving the human-level AI society.
Concluding remarks
Having undergone twenty years of development from 2000 to now, Web Intelligence has grown into one of the most popular research felds. As an enhancement or an extension of AI and IT, its initial goal was to develop wisdom Web centric products, systems, services and applications to meet the requirements of the rise of knowledge economies. For the next twenty years, Web Intelligence will be on an ever-greater role in the connected world towards the organic amalgamation and harmonious symbiosis among humans, computers and things. In this paper, we have presented the systematic review of Web Intelligencerelated topics with respect to people, data, things, trust and agents, respectively. Furthermore, the role of “brain intelligence” and “intelligence technology” is discussed to promote the development of Web Intelligence, specifcally from the viewpoint of “Web Intelligence meets Brain Informatics”. Finally, we highlight two representative directions that are Data-Brain oriented research and brain-inspired wisdom service to accelerate the practice of Web Intelligence and Brain Informatics. More and more investigators will contribute their knowledge and experience in areas concerning the new paradigm of “Web Intelligence meets Brain Informatics”, which will produce new efciencies and provide a strong catalyst towards achieving the ultimate goal of a human-level AI society.