论科学发现作为问题解决的特殊性——对赫伯特·西蒙机器发现研究的分析

将科学发现视为问题解决的观点,是科学哲学与认知研究相结合的产物,促成机器发现研究方法的产生。西蒙不仅指出科学发现作为问题解决的普遍特征,还概括了科学发现区别于其他问题解决的个性特征。这些个性特征从创造性的角度将科学发现与常规科学的问题解决区分开来。从科学问题的角度看,科学发现和机器发现都是从未知到已知的科学问题解决过程,体现了不同层面的创造性,使得机器发现可以通过再发现的形式模拟科学发现。     

Creating a discoverer: Autonomous knowledge seeking agent

Construction of a robot discoverer can be treated as the ultimate success of automated discovery. In order to build such an agent we must understand algorithmic details of the discovery processes and the representation of scientific knowledge needed to support the automation. To understand the discovery process we must build automated systems. This paper investigates the anatomy of a robot-discoverer, examining various components developed and refined to a various degree over two decades. We also clarify the notion of autonomy of an artificial agent, and we discuss the ways in which machine discoverers become more autonomous. Finally we summarize the main principles useful in construction of automated discoverers and we discuss various possible limitations of automation.     

高平开化寺“立机子”图像探微

高平开化寺壁画有"立机子"织造图一幅,成为宋代踏板单综立式织机目前仅见遗存。其形制与金末元初薛景石《梓人遗制》中所载"立机子"相近,两幅图像具有高超的写实绘画技术,其真实性较高。从织机结构与装配工艺分析,该织机与薛氏"立机子"在中轴及马头等诸多零件上不尽相同,其分经工艺差异性较大。这两幅图像也呈现出晋南、晋东南家庭作坊式织造业发展过程中区域工具特征,是宋元时期北方小型织机代表,有着极高的图像史料价值。     

Recent Advances in Predictive (Machine) Learning

Prediction involves estimating the unknown value of an attribute of a system under study given the values of other measured attributes. In prediction (machine) learning the prediction rule is derived from data consisting of previously solved cases. Most methods for predictive learning were originated many years ago at the dawn of the computer age. Recently two new techniques have emerged that have revitalized the field. These are support vector machines and boosted decision trees. This paper provides an introduction to these two new methods tracing their respective ancestral roots to standard kernel methods and ordinary decision trees.     

科学发现的本质及其逻辑机制的再发现

不管对科学发现进行语义分析，还是对科学发现活动本身进行历史探究，我们都会发现，科学发现除了具有社会、历史和心理上的属性外，它在本质上乃是新假说的产生、选择、修正、接受和解释的逻辑过程，其逻辑机制正是皮尔士与汉森等所倡导的溯因推理。而且，现代认知科学的新成果促进了溯因推理形式的不断完善，人们意识到在新假说产生与选择中背景理论与背景知识的重要作用，创新了溯因推理的推理形式，令人满意地回答了溯因推理何以能够以及如何能够成为科学发现的逻辑的问题。     

图林的智能机器思想

本文讨论了提出著名“理想计算机”理论的英国数学家阿兰.图林半个世纪前在机器智能研究领域的重要贡献,证明他的机器智能思想包含了关于机器智能可能性的论证、机器智能水平的判据(图林测验)、包含学习机器在内的广义计算机分类体系等内容,并探讨了图林智能计算机思想与维纳控制论思想的相互影响。     

基于自注意力机制的科技术语自动提取技术研究

科技术语提取是科技术语自动处理的重要环节,对后续的机器翻译、信息检索、QA问答等任务有重要意义.传统的人工科技术语提取方法耗费大量的人力成本.而一种自动提取科技术语方法是将术语提取转化为序列标注问题,通过监督学习方法训练出标注模型,但是面临缺乏大规模科技术语标注语料库的问题.文章引入远程监督的方法来产生大规模训练标注语...     

论古代科学发现优先权的确立原则与判定标准——以勾股定理与毕达哥拉斯定理的发现为例

通过对勾股定理与毕达哥拉斯定理发现与证明方法比较,考证了中西方对勾股定理的发现均符合现代科学发现的定义,提出了中国与西方几乎同时独立发现勾股定理。根据时间敏感性差异原理,指出了古代科学发现优先权的确立原则和判定标准,中西方这种独立的发现开辟了中西方科学发展的不同模式。     

Discovering the Principle of Finality in Computational Machines

In this essay we argue that the notion of machine necessarily includes its being designed for a purpose. Therefore, being a mechanical system is not enough for being a machine. Since the experimental scientific method excludes any consideration of finality on methodological grounds, it is then also insufficient to fully understand what machines are. Instead in order to understand a machine it is first required to understand its purpose, along with its structure, in clear parallel with Aristotle’s final and formal causes. Obviously, purpose and structure are not machine components that can physically interact with other components; nonetheless they are essential to understanding their operation. This casts an interesting light on the relationship between mind and body: for just as an artifact’s finality and structure explain its operation, so also consciousness is the explanation—not the efficient cause—of specifically human behavior. What machines and human beings have in common is that, in order to understand them, it is necessary to appeal to the principle of finality. Yet while finality is given and extrinsic in the case of machines, we human beings are characterized by the ability to self-propose our own ends. Since the principle of finality is essential to understanding the production of machines, the traditional view in modern Western philosophy that finality lies beyond the scope of objective/scientific knowledge should be rectified to allow for a genuine science of the artificial. We think a correct understanding of final causality will overcome current resistance to this principle.     

Rationality in the Discovery of Empirical Laws

In this paper I argue against the traditional viewthat in discovery processes there is no place forrational decisions. First I argue that some historicalprocesses in which an empirical law was developed,were rational. Second, I identify some of themethodological rules that we can follow in order to berational when constructing an empirical law. Finally,I argue that people who deny that scientific discoverycan be rational do not understand the nature ofmethodological rules.     

基于多模态术语知识库的译者知识获取

掌握翻译实践所涉及的专业领域知识以及获取相关知识的方法,是译者应具备的职业素养,也是提升翻译服务质量的必然要求。多模态术语知识库整合语料库、术语库、关系库等语言知识素材,建立面向译者需求的数据分析机制与知识获取机制,并以可视化的人机交互手段优化知识的表示与利用环节,降低专业领域的知识壁垒。基于此,译者能够从中高效地获取和习得翻译过程中不可或缺的语言类与专业类知识,包括术语知识、搭配知识、概念实体知识与逻辑关系知识。     

Can Darwinian Mechanisms Make Novel Discoveries?: Learning from discoveries made by evolving neural networks

Some philosophers suggest that the development of scientificknowledge is a kind of Darwinian process. The process of discovery,however, is one problematic element of this analogy. I compare HerbertSimon's attempt to simulate scientific discovery in a computer programto recent connectionist models that were not designed for that purpose,but which provide useful cases to help evaluate this aspect of theanalogy. In contrast to the classic A.I. approach Simon used, ``neuralnetworks' contain no explicit protocols, but are generic learningsystems built on the model of the interconnections of neurons in thebrain. I describe two cases that take the connectionist approach a stepfurther by using genetic algorithms, a form of evolutionary computationthat explicitly models Darwinian mechanisms. These cases show thatDarwinian mechanisms can make novel discoveries of complex, previouslyunknown patterns. With some caveats, they lend support to evolutionaryepistemology.     

牛顿科学的"推理规则"与科学发现的逻辑

牛顿在<自然哲学的数学原理>中提出的"哲学中的推理规则",不仅浓缩了牛顿时代有关科学方法论的许多内容,而且还蕴涵着在此之后有关"科学发现的逻辑"问题的历史变化的思想根源,甚至这一规则也可以作为今天"复活发现逻辑"的一个典型的科学史范例.     

人工智能“合乎伦理设计”的理论探源

人工智能设计是为了满足设计主体的目的而制造人工物的过程,设计是合目的性与合价值性的统一,设计体现设计者的意愿,设计关注设计的目的、行为以及人工物的功能。与其他技术人工物不同,智能机器具有自身推理能力与自动决策功能,为了更好地设计人工智能客体,为了充分发挥人工智能的功能,必须对其理论基础追踪溯源,对其理论来源与理论前提进行充分挖掘,才能保证人工智能设计在设计之初把人类规范与道德嵌入到智能机器中,才能规范智能机器的发展。     

Reasoning to Hypotheses: Where Do Questions Come?

Detectives and scientists are in the business of reasoning from observations to explanations. This they often do by raising cunning questionsduring their inquiries. But to substantiate this claim we need to know how questions arise and how they are nurtured into more specific hypotheses. I shall discuss what the problem is, and then introduce the so-called interrogative model of inquiry which makes use of an explicit logic of questions. On this view, a discovery processes can be represented as a model-based game in which an inquirer subjects a source of information to a series of strategically organized questions. Strategic principles and why-questions are especially important in heuristical reasoning. Why-questions have their own peculiar nature among questions. They indicate that the inquirer's expectations are somehow disappointed, and that is cognitively challenging. In a finished argument why-questions can be omitted, but in the search for more specific questions they are highly important. As a detetective example I shall analyze Sherlock Holmes reasoning in Silver Blace, the scientific one is A.R. Wallace's discovery of the principle of natural selection. In both of these examples the meaning of questions, especially of well-chosen why-questions, of strategic principles, and of highly structured background knowledge come to the fore. Good questions frequent those who have orderly expectations, based on experience and expertise (detectives!) or highly structured background theories (scientists!).     

Inference to the Best Manipulation – A Case Study of Qualitative Reasoning in Neuropharmacy

How can new drug lead suggestions beinferred from neurophysiological models? This paperaddresses this question based on a case study ofresearch into Parkinson's disease at the GroningenUniversity Department of Pharmacy. It is argued thatneurophysiological box-and-arrow models can beunderstood as qualitative differential equationmodels. An inference task is defined to helpunderstand and possibly aid the discovery andexplanation of new drug lead suggestions.     

塞尔"中文屋"思想实验的哲学意蕴辨析

本文认为塞尔“中文屋”思想实验的要点是提出了意向性是人机之间的区别。但根据塞尔本人对意向性的定义，是以人的主观状态为前提的，因而不能成为人机之间的区别。而计算机只不过是西方理性文化高度发展的结果，是形式化思维方式的产物。     

科技史研究中的"双重标准"现象

科技史研究中存在着对同类问题的判定采用“双重标准”的现象。同一作者的“双重标准”是逻辑欠严密,首尾不一贯;不同作者之间存在的“双重标准”现象则是学术观点不同的结果。“双重标准”不利于科学发现优先权的评价,需要通过学术讨论统一认识,这至标准统一。     

Abduction as a Logic and Methodology of Discovery: the Importance of Strategies

There are various ``classical' argumentsagainst abduction as a logic of discovery,especially that (1) abduction is too weak amode of inference to be of any use, and (2) inbasic formulation of abduction the hypothesisis already presupposed to be known, so it isnot the way hypotheses are discovered in thefirst place. In this paper I argue, bybringing forth the idea of strategies,that these counter-arguments are weaker thanmay appear. The concept of strategiessuggests, inter alia, that many inferentialmoves are taken into account at the same time.This is especially important in abductivereasoning, which is basically a very weak modeof inference. The importance of strategicthinking can already be seen in Charles S.Peirce's early treatments of the topic, and N.R.Hanson's later writings on abductionalthough they did not use the concept of``strategies.' On the whole, I am arguing thatthe focus should be more on methodologicalprocesses, and not only on validityconsiderations, which have dominated thediscussion about abduction.