The red thread of the AI-IP-EI Trilogy fate of this study, may have the appearance of a pot-pourri of intellectual and intelligence natures, as a matter of fact that it emanates from the genesis and practical synergis...The red thread of the AI-IP-EI Trilogy fate of this study, may have the appearance of a pot-pourri of intellectual and intelligence natures, as a matter of fact that it emanates from the genesis and practical synergism of the trilogy components. Concretely: The paper goes from: AI (Artificial Intelligence)—to the related IP (Intellectual Property) domain—to the relevance of EI (Emotional Intelligence);thus, forming the new AI-IP-EI Trilogy and its attributes and specific impacts to the new innovation process, and business model dimensions. These impacts are outlined and illustrated in part in essays of specific sections and all along. Several concrete study cases are used in the various dedicated sections;such as cases respective to the inventor status, and the EI factor, to the sport education innovative dimension, as well as to biases as inevitably promoting and revealing, to drastically enlarge open innovation supported by constructivism and creations of musical group as a model of open reflexive education. Overall resulting adapted business models appear to have a massive potential, and a multidimensional reach with a necessary attention to the IP policy on going definition. The durable green dimension is exemplified as well. The Ethics-plus, “@LEAST<span style="font-family:Verdana, Helvetica, Arial;white-space:normal;background-color:#FFFFFF;">?</span>”, said corpus is proposed. Could a human centric, AI-adapted-IP policy, internationally embraced, take part to some level of arbitrage, normative and enduring reliability in the field of interest? This seems to be “en route”. Shall the EI (Emotional Intelligence) factor be supervised? Likely so. Is traditional open innovation renewed to a more comprehensive, more inclusive dimension reminding best business practice and now “beyond”? Definitely, and will remain an opportunity, all along the 4IR quantum game changer to come. Neither seeking an in-depth expert analysis, nor a grand public over-simplified bavardage, of the trilogy, AI-IP-EI, four authors here propose an illustrated view of scientific, educational, visionary, demonstrative value to the subject matter. They are aged about 30-40-50-60, being IP & Innovation strategist, future IP lawyer, children-teacher and professional academy sport coach, illustrator and bio-advanced materials engineering “Fellow Scientist”. With experience of large and smaller organizations, being involved innovators, inventors and private artists as well, they are sharing their “non-jargonized down-to-earth”, forward looking views through a structured analysis of the trilogy using realistic examples and data from rather diverse specialized independent sources, biotechnology, nanomaterials, sport… New invention and inventorship is been “reconceptualized” at least from an “insighter or insider” viewpoint, and sport team approach more broadly revisited from its academy level to its commercial asset impact, via educational virtues and values. Music group constructivism enters the scene as well with its exemplary reflexivity and alterity valued for open innovation. Science is the prime lead. “Emotional intelligence, EI, is still an emerging area within AI” and beyond? A new open innovation scheme is taking place. This prompted our intention to further contribute to this matter. Is EI, the tree gently challenging the wind? Generated by AI and IP streams and scientific applications therewith? Naturally. Conclusions are encouraging the follow-up of promising orientations underlined by the AI-IP-EI Trilogy, favoring human centric feature adoptions.展开更多
为充分挖掘专利文本中已有的解决方案和技术知识,依据发明问题解决理论(theory of inventive problem solving,TRIZ),提出了一种基于预训练语言模型的方法,将其用于面向TRIZ发明原理的中文专利分类研究中。基于整词掩码技术,使用不同数...为充分挖掘专利文本中已有的解决方案和技术知识,依据发明问题解决理论(theory of inventive problem solving,TRIZ),提出了一种基于预训练语言模型的方法,将其用于面向TRIZ发明原理的中文专利分类研究中。基于整词掩码技术,使用不同数量的专利数据集(标题和摘要)对中文RoBERTa模型进一步预训练,生成特定于专利领域的RoBERTa_patent1.0和RoBERTa_patent2.0两个模型,并在此基础上添加全连接层,构建了基于RoBERTa、RoBERTa_patent1.0和RoBERTa_patent2.0的三个专利分类模型。然后使用构建的基于TRIZ发明原理的专利数据集对以上三个分类模型进行训练和测试。实验结果表明,RoBERTa_patent2.0_IP具有更高的准确率、宏查准率、宏查全率和宏F 1值,分别达到96%、95.69%、94%和94.84%,实现了基于TRIZ发明原理的中文专利文本自动分类,可以帮助设计者理解与应用TRIZ发明原理,实现产品的创新设计。展开更多
文摘The red thread of the AI-IP-EI Trilogy fate of this study, may have the appearance of a pot-pourri of intellectual and intelligence natures, as a matter of fact that it emanates from the genesis and practical synergism of the trilogy components. Concretely: The paper goes from: AI (Artificial Intelligence)—to the related IP (Intellectual Property) domain—to the relevance of EI (Emotional Intelligence);thus, forming the new AI-IP-EI Trilogy and its attributes and specific impacts to the new innovation process, and business model dimensions. These impacts are outlined and illustrated in part in essays of specific sections and all along. Several concrete study cases are used in the various dedicated sections;such as cases respective to the inventor status, and the EI factor, to the sport education innovative dimension, as well as to biases as inevitably promoting and revealing, to drastically enlarge open innovation supported by constructivism and creations of musical group as a model of open reflexive education. Overall resulting adapted business models appear to have a massive potential, and a multidimensional reach with a necessary attention to the IP policy on going definition. The durable green dimension is exemplified as well. The Ethics-plus, “@LEAST<span style="font-family:Verdana, Helvetica, Arial;white-space:normal;background-color:#FFFFFF;">?</span>”, said corpus is proposed. Could a human centric, AI-adapted-IP policy, internationally embraced, take part to some level of arbitrage, normative and enduring reliability in the field of interest? This seems to be “en route”. Shall the EI (Emotional Intelligence) factor be supervised? Likely so. Is traditional open innovation renewed to a more comprehensive, more inclusive dimension reminding best business practice and now “beyond”? Definitely, and will remain an opportunity, all along the 4IR quantum game changer to come. Neither seeking an in-depth expert analysis, nor a grand public over-simplified bavardage, of the trilogy, AI-IP-EI, four authors here propose an illustrated view of scientific, educational, visionary, demonstrative value to the subject matter. They are aged about 30-40-50-60, being IP & Innovation strategist, future IP lawyer, children-teacher and professional academy sport coach, illustrator and bio-advanced materials engineering “Fellow Scientist”. With experience of large and smaller organizations, being involved innovators, inventors and private artists as well, they are sharing their “non-jargonized down-to-earth”, forward looking views through a structured analysis of the trilogy using realistic examples and data from rather diverse specialized independent sources, biotechnology, nanomaterials, sport… New invention and inventorship is been “reconceptualized” at least from an “insighter or insider” viewpoint, and sport team approach more broadly revisited from its academy level to its commercial asset impact, via educational virtues and values. Music group constructivism enters the scene as well with its exemplary reflexivity and alterity valued for open innovation. Science is the prime lead. “Emotional intelligence, EI, is still an emerging area within AI” and beyond? A new open innovation scheme is taking place. This prompted our intention to further contribute to this matter. Is EI, the tree gently challenging the wind? Generated by AI and IP streams and scientific applications therewith? Naturally. Conclusions are encouraging the follow-up of promising orientations underlined by the AI-IP-EI Trilogy, favoring human centric feature adoptions.
文摘为充分挖掘专利文本中已有的解决方案和技术知识,依据发明问题解决理论(theory of inventive problem solving,TRIZ),提出了一种基于预训练语言模型的方法,将其用于面向TRIZ发明原理的中文专利分类研究中。基于整词掩码技术,使用不同数量的专利数据集(标题和摘要)对中文RoBERTa模型进一步预训练,生成特定于专利领域的RoBERTa_patent1.0和RoBERTa_patent2.0两个模型,并在此基础上添加全连接层,构建了基于RoBERTa、RoBERTa_patent1.0和RoBERTa_patent2.0的三个专利分类模型。然后使用构建的基于TRIZ发明原理的专利数据集对以上三个分类模型进行训练和测试。实验结果表明,RoBERTa_patent2.0_IP具有更高的准确率、宏查准率、宏查全率和宏F 1值,分别达到96%、95.69%、94%和94.84%,实现了基于TRIZ发明原理的中文专利文本自动分类,可以帮助设计者理解与应用TRIZ发明原理,实现产品的创新设计。
