原站英文机器转录 + 本地中文机器初译(未完成逐句听审) 原作者及许可见 attribution.txt;中文由九天新增,不代表原作者认可。 00:00:10.930 EN: Welcome to the Knowledge Engineering and Semantic Web lecture. Last week, you were discussing with Markus RDF schema or starting to discuss. 中文初译: 欢迎来到知识工程和语义网站讲座。 上周, 你和Markus“ RDF” Schema 讨论或开始讨论。 00:00:22.710 EN: So let's maybe quickly recap what has been there about RDF schema. And if we look at the linked data stack or the semantic web layer cake, RDF schema builds on top of RDF. 中文初译: 所以,让我们快速回顾一下“RDF”的图案。如果我们看看连接的数据堆或语义网络层蛋糕, “RDF”的图案就建在了“RDF”之上。 00:00:39.650 EN: So basically, RDF schema is just some additions to RDF. The data model itself is not changed or altered by RDF schema. 中文初译: 基本上,“RDF”的图案只是对“RDF”的一些补充。数据模型本身没有被“RDF”的图案改变或修改。 00:00:53.760 EN: And you all know RDF Meanwhile and the RDF data model. With RDF, we can express facts like Albert Einstein discovered theory of relativity, for example. 中文初译: 你们都知道“RDF”同时和“RDF”数据模型。用“RDF”,我们可以表达事实,比如艾伯特·爱因斯坦发现了相对论。 00:01:09.880 EN: But how can we refine such facts? How can we define that the predicate discovered has a person as a subject and the theory as an object? 中文初译: 但是,我们如何改进这些事实呢?我们如何界定被发现的上游人将一个人作为主体和理论作为对象? 00:01:20.940 EN: Because you can imagine, we could create all kinds of weird statements. So RDF itself doesn't constrain the statements in any way. 中文初译: 因为你们可以想象,我们可以创造各种怪异的言论。所以“RDF”本身不会以任何方式限制这些言论。 00:01:35.520 EN: So you could also create pretty strange, weird statements. And that's what RDF schema is about constraining RDF in such a way that knowledge bases make more sense. 中文初译: 所以,你也可以创造出非常奇怪、怪异的语句。这就是“RDF”的策略 限制“RDF”的方式, 使知识基础更有意义。 00:01:55.680 EN: So how can we express that about Einstein was a researcher and that every researcher is a human? And such knowledge is called schema knowledge or also terminological knowledge. 中文初译: 那么,我们如何表达爱因斯坦是一个研究人员,每个研究人员都是人类?这种知识被称为“假想知识 ” , 或者也称为术语知识。 00:02:07.540 EN: And that's what RDF schema is about. So there is an W3C standard, RDFS. In short, RDF schema, it's part of the W3C recommendation family. 中文初译: 这就是“RDF”计划的目的。所以有一个W3C标准,“RDFS”。简言之,“RDF”计划,它是W3C推荐家庭的一部分。 00:02:22.990 EN: It's used for such schema and terminological knowledge. And it's itself an RDF vocabulary. So every RDF schema graph is also an RDF graph. 中文初译: 它用来做这样的图案和术语学知识。它本身就是一个“RDF”词汇。 所以每个“RDF”的图案图也是“RDF”图表。 00:02:38.700 EN: The vocabulary is generic. So it's not bound to a specific application area. And it allows to specify domain specific or user defined RDF vocabularies. 中文初译: 词汇表是通用的。 所以它不受特定应用区域的约束。 它允许指定特定的域或用户定义的“ RDF” 词汇表 。 00:02:50.580 EN: And there is a namespace, which you can see here, like URI namespace. And the typical prefix is RDFS. So the first thing that you can do in RDF schema is defining classes. 中文初译: 这里有一个命名空间, 您可以在这里看到, 比如“ URI” 命名空间。 典型的前缀是“ RDFS ” 。 所以在“ RDF” 模式中你可以做的第一件事就是定义类 。 00:03:05.900 EN: A class is a set of things or entities. And these are identified by URIs. You can also imagine classes as sets. 中文初译: A类是一系列东西或实体。这些是由“URIs”识别的。您也可以想象类是一组。 00:03:14.780 EN: And the membership of an entity to a class is defined by using the RDF type property. So for example, the fact that my blue faraway golf is a member instance of the class XCAR can be expressed by the statement my blue faraway golf RDF type XCAR. 中文初译: 比如,我的蓝色远方高尔夫是XCAR级成员的例子,这可以通过我所说的“RDF”型的蓝色远方高尔夫(RDF)高尔夫(RDF)高尔夫(RDF)高尔夫(RDF ) 。 00:03:36.520 EN: So the instance on a concrete car belongs to the class car. And as I said, you can imagine the class as a set of instances. 中文初译: 所以混凝土车的例子属于类车。正如我所说, 你可以想象出这门课是一系列例子。 00:03:48.140 EN: And also resource can belong to several classes. So it can belong to the class car, but also to the class, for example, German product, as you can see here in the lower example. 中文初译: 资源也可以属于几个类。因此它可以属于类汽车, 但也属于类产品,例如德国产品, 从下面的例子中可以看到。 00:04:01.400 EN: We can then build hierarchies of classes. We can say that car is a subclass of motor vehicle. And that would then mean that my concrete golf car is also a motor vehicle. 中文初译: 然后我们可以建立班级等级。我们可以说汽车是机动车辆的亚类。那将意味着我的水泥高尔夫车也是机动车辆。 00:04:14.000 EN: And that's what we often have in reality in the nature. We have often these kind of subclass relationships. Since we can imagine classes or picture classes as sets, the subclass of relationship basically is then a subset relationship. 中文初译: 这就是我们通常在自然界中的现实。我们经常有这种子类关系。由于我们可以想象成类类或图片类,那么子类关系基本上就是一种子类关系。 00:04:35.080 EN: So the set of cars is a subset of the set of motor vehicles. That means every car is also a motor vehicle, but not the other way around. 中文初译: 所以汽车组是机动车辆组的子集。这意味着每辆汽车也是机动车辆,而不是相反。 00:04:44.960 EN: A motor vehicle doesn't have to be a car. There are also other motor vehicles. You can imagine motorbikes, trucks. They are also motor vehicles. 中文初译: 机动车辆不必是汽车,还有其他机动车辆。你可以想象摩托车和卡车。它们也是机动车辆。 00:04:58.000 EN: And this is already the first example of implicit knowledge. And that distinguishes RDF and RDF schema from other data modeling techniques. 中文初译: 这已经是隐含知识的第一个例子。这把“RDF”和“RDF”模型与其他数据建模技术区分开来。 00:05:09.580 EN: And usually in other data modeling techniques, like in relational databases or in XML, we don't have implicit knowledge. 中文初译: 通常在其他数据模型技术中, 比如关系数据库或XML, 我们没有隐含的知识。 00:05:19.680 EN: In RDF and RDF schema, we have this implicit knowledge. So we can infer that my blue faraway golf is also a motor vehicle without this statement being explicitly stated in the knowledge base. 中文初译: 在“RDF ” 和“RDF ” 计划中,我们有这种隐含的知识。所以我们可以推断我的蓝色远方高尔夫球也是一部机动车,而知识库却没有明确说明这一说法。 00:05:34.620 EN: So you see the upper two statements, triples, are in the knowledge base. And we can infer a third triple from the knowledge base resulting from the subclass of relationship. 中文初译: 所以,你可以看到上两个对白,三倍, 都在知识库中。我们可以推断出三倍于 由关系子类产生的知识库。 00:05:49.880 EN: So that's revealing implicit knowledge. And if we add also that motor vehicle is a subclass of vehicle, another inference can be done. 中文初译: 这就是隐含的知识。如果我们加上机动车是车辆的子类, 就可以做出另一种推论。 00:06:08.910 EN: We can infer that car is also a subclass of vehicle because the subclass of relationship is transitive. Now, I think you have all heard about transitive relationships and math. 中文初译: 我们可以推断那辆车也是个小车类 因为这小关系类是过渡性的。现在,我想你们都听说过 过渡关系和数学。 00:06:22.550 EN: So this holds here for the subclass of relationship. We can define our own classes. So we just say that our own class is of type RDFS class. 中文初译: 因此,这保留在关系子类中。我们可以定义我们自己的班级。 所以我们只说我们自己的班级属于“RDFS”类。 00:06:42.120 EN: So that's what you see here. Our car class, we associated the type RDFS class. And with that, we make car basically an instance of RDFS class. 中文初译: 这就是你在这里看到的。我们的汽车类,我们把“RDFS”类联系起来。因此,我们把汽车做成一个“RDFS”类的例子。 00:06:55.840 EN: And that makes it a class. And of course, we can also instantiate then the car class. So that's interesting that we use RDF type on the one hand to define classes, but on the other hand, also to assign instances to classes. 中文初译: 这使得它成为一个类。当然,我们也可以即时进行汽车类。所以有趣的是,我们一方面使用“RDF”类型来定义班级,另一方面也给班级分配实例。 00:07:13.060 EN: So my blue faraway golf can have the type car. And then we can associate it with the class car. So this makes RDFS class also itself an instance of RDFS class. 中文初译: 所以,我的蓝色远方高尔夫可以拥有车型。然后我们可以把它和类车联系起来。所以这让“RDFS”类本身也成为了“RDFS”类的例子。 00:07:31.380 EN: So RDFS class is of type RDFS class. So we can also then define the equivalence of classes. And it happens often that people are independently from each other or organizations defining classes. 中文初译: 所以“ RDFS” 类属于“ RDFS ” 类。 所以我们也可以定义各类的等同性。 经常发生的情况是,人们彼此独立, 或由不同的组织来定义各类。 00:07:53.360 EN: And later on, you discover that they actually mean the same. Of course, you should avoid if you create a vocabulary and ontology, you should avoid defining two classes which mean the same. 中文初译: 稍后,你发现它们的意思其实是相同的。当然,如果你创建了词汇和本体论,你应该避免定义两个意思相同的类别。 00:08:06.320 EN: But since the semantic web is as the web a decentralized information space, everybody can define his own vocabularies and schemas. 中文初译: 但既然语义网站是作为网络的分散信息空间, 每个人都可以定义自己的词汇和计划。 00:08:19.060 EN: And that's why it's quite likely that there might be different classes appearing on the internet being defined in vocabularies, ontologies which actually mean the same. 中文初译: 这就是为什么在互联网上 很可能存在不同的类别 在词汇中被定义, 理论学实际上意味着相同的。 00:08:30.480 EN: So and we can then express equivalence of two classes. We can basically establish a connection between previously two independent classes and say that they are meaning the same. 中文初译: 因此,我们可以表达两种等级的等同性。我们基本上可以在之前的两个独立的类别之间建立一种联系,并说它们的含义是一样的。 00:08:42.280 EN: And that can be done by making them mutually subclasses of each other. So we can, for example, say that car is a subclass of automobile and automobile is a subclass of car. What does it mean? 中文初译: 这可以通过使它们相互相联的子类来做到。例如,我们可以说汽车是汽车的子类,汽车是汽车的子类。这是什么意思? 00:08:56.240 EN: It basically means that the set of cars is a subset of the set of automobiles. That's the first statement. The second statement says the set of automobiles is a subset of the set of cars. 中文初译: 这基本上意味着这组汽车是这组汽车的子集。这是第一组声明。第二组声明说这组汽车是那组汽车的子集。 00:09:09.800 EN: And you can already figure out that when you have this mutual subset relationship, the sets must be actually equal. 中文初译: 并且你已经可以发现 当你有了这种共同的子集关系时, 这些组合必须实际上是相等的。 00:09:18.840 EN: So the set of cars is the same as the set of automobiles. And that means the two cars basically are the same. And that leads also to the statement that car is also a subclass of car. 中文初译: 所以汽车组和汽车组是相同的。这意味着两辆汽车基本上是一样的。这也导致一种说法,即汽车也是一辆小汽车。 00:09:35.400 EN: So the subclass of relationship is not only transitive but it's also reflexive. Each set is also a subset of itself, not a proper subset but a subset. 中文初译: 所以关系子分类不仅具有过渡性,而且具有反射性。每组都本身也是一个子集,不是适当的子集,而是子集。 00:09:48.600 EN: And that's why RDFS subclass of is a reflexive relationship. So there are some predefined RDFS classes in the RDFS vocabulary. 中文初译: 这就是为什么“RDFS”子类是一种反射关系。所以在“RDFS”词汇中有一些预定义的“RDFS”类。 00:10:00.830 EN: So for example, RDFS resource is the class of all things. It's the superclass of all classes. So everything basically belongs to RDFS resource. 中文初译: 例如,“RDFS”资源是所有事物的类。它是所有阶级的超级类。所以所有的东西基本上都属于“RDFS”资源。 00:10:13.070 EN: Then we have RDF property. That's the class of all properties. Properties are those resources which are used at the predicate position. 中文初译: 然后我们拥有“RDF”财产。这是所有财产的类别。 属性是那些在上游位置使用的资源。 00:10:22.230 EN: We have subject predicate object, right? At the middle position of our statements. And these are the properties and they belong to the class of RDF property. 中文初译: 我们的口供处于中间位置, 这些是属于“卢旺达国防部队”财产的属性。 00:10:34.650 EN: Then we have the class RDFS data type. That's the class of all data types. So every instance of this class is a subclass of RDFS literal. 中文初译: 然后我们有了“RDFS”类数据类型。 这是所有数据类型的类别。 因此,这个类别的每一个例子都是“RDFS”字面上的子类。 00:10:51.530 EN: So these are basically the comprise all the literals which can appear at the object position. And RDFS literal is the class of literal values such as strings or integers. 中文初译: 所以这些基本上是包含所有可以出现在对象位置的字面内容。 而“RDFS”字面是字符串或整数等字面值的类别。 00:11:14.430 EN: And if a literal is not typed then it is an instance of RDFS data type. Then we have some more predefined RDFS classes, RDFS. 中文初译: 如果字面上没有输入, 那么它就是“ RDFS” 数据类型的例子。 然后我们有一些更预定义的“ RDFS” 类, “ RDFS ” 。 00:11:30.360 EN: RDFLangString is the class of language text string literals which is an instance of RDFS data type and the subclass of RDFS literal. 中文初译: RDFLangString 是语言文字字符串字典的类别, 它是一个“ RDFS” 数据类型的实例, 以及“ RDFS” 字典的子类 。 00:11:41.850 EN: So these are these language strings like the E or EN for English which you can use to tag the language of a string literal. 中文初译: 这些语言字符串就像英语的 E 或 EN, 您可以用来标记字符串字典的语言 。 00:11:54.250 EN: Then there is the class RDFXML literal which is the class of XML literal values. So you can also have literals which contain XML snippets, yeah? 中文初译: 然后是类 RDFXML 字典, 也就是“ XML” 字义值。 这样您也可以有包含“ XML” 片段的字典, 是么 ? 00:12:06.690 EN: And these are then XML literals. So it's a subclass of RDFS literals and also an instance of RDFS data type. It's a specific data type. 中文初译: 然后这些是“XML”字的字面。所以它是一个“RDFS”字面的子类, 和一个“RDFS”数据类型的例子。 它是一个特定的数据类型。 00:12:21.190 EN: And there are also more data types we have seen that we can use RDFXML schema data types. So they are defined in the XML standard for example, for dates, for numbers you have predefined data types and they can be reused in RDF and RDF schema of course and they are then also belong to the class RDFS data type. 中文初译: 我们还看到更多的数据类型, 我们可以使用 RDFXML schema 数据类型。 因此在“ XML” 标准中定义了这些数据类型, 比如日期, 您预设的数据类型, 可以被重新用于“ RDF” 和“ RDF” 模式, 然后它们也属于“ RDFS” 类数据类型 。 00:12:46.050 EN: Then we have a class RDFS, RDF statement. So you already see here, sometimes we have RDFS sometimes we have RDF. 中文初译: 然后我们有一个“RDFS”、“RDF”的类声明。你在这里已经看到,有时我们有“RDFs” 有时我们有“RDF”。 00:12:56.010 EN: So in the beginning when RDF and RDF schema were defined this distinction was maybe not from the beginning very clear first RDF was defined later on RDF schema was added. 中文初译: 所以在“RDF”和“RDF”模式被定义的一开始, 可能从一开始这个区别就不是非常清楚。 “RDF”后来在“RDF”模式上被添加了。 00:13:08.830 EN: That's why some of these classes were already defined in RDF. Some were then added in RDF schema. That's why RDF statement is something which was already defined in the RDF standard. 中文初译: 这就是为什么其中一些类已经在“RDF”中定义了。有些则被添加到“RDF”的模型中。这就是为什么“RDF”的语句已经定义在“RDF”标准中的原因。 00:13:24.690 EN: And this is the class of all RDF statements. Statements are the subject predicate object triples. So every RDF triple can be seen as an instance of this class with a RDF subject, RDF predicate and RDF object property. 中文初译: 这是所有“ RDF ” 语句的类别。 语句是标的上游物体三倍。 所以每个“ RDF ” 三倍都可以被视为该类中的一例, 其中含有“ RDF ” 主题、 “ RDF ” 上游和“ RDF ” 对象属性。 00:13:43.130 EN: So we will see this later on. This is also called reification. I will show you some examples how that works. And then we have containers. 中文初译: 稍后我们再看这个。这也称为“重新化”。我将向大家展示一些例子,说明它是如何运作的。然后我们有容器。 00:13:54.030 EN: So RDFS container is a superclass for container classes RDF back sequence and alternative. And I mentioned already that some of these here are indeed already in the part of the RDF vocabulary. 中文初译: 所以“RDFS”容器是集装箱级“RDF”背序和替代的超级级。我已经提到过,其中一些已经进入了“RDF”词汇中。 00:14:12.510 EN: So that's why they have the namespace prefix RDF. And in RDF schema, they are declared to be classes. They were already available before as resources. 中文初译: 这就是为什么他们有“RDF”的命名空间前缀。在“RDF”模式中,他们被宣布为班级。它们以前已经作为资源可用了。 00:14:23.470 EN: And in RDF schema, they are defined to be classes as well. So we can then also define own properties as we can define classes, we can define new properties. 中文初译: 在“RDF”模式中,它们也被定义为类。 所以我们也可以定义自己的属性, 因为我们可以定义类别, 我们可以定义新的属性。 00:14:37.370 EN: For example, if we want to express that there is a new property, we define it as an instance of the property class. 中文初译: 例如,如果我们想表示有新财产,我们就把它定义为财产类别的例子。 00:14:47.490 EN: For example, drives is of type RDF property. And with this new property, we can express that max drives a Volkswagen Golf, not just any one, but a specific one. 中文初译: 例如,驱动器是“RDF”类型的属性。有了这种新的属性,我们可以表示最大驱动器是大众高尔夫, 不仅仅是任何一种,而是特定的。 00:15:01.290 EN: We can then say max drives my blue Volkswagen Golf, for example. You don't have to define properties. You can also just use a resource as a property, then it automatically implicitly becomes a property, right? 中文初译: 然后我们可以说"最大"驱动我的蓝色大众高尔夫,例如。你不必定义属性。你也可以将资源作为财产使用,然后自动暗含成为财产,对吗? 00:15:16.810 EN: If you just use drives as a property, as a predicate in a statement, then implicitly you can infer that what's at the predicate position in a triple belongs to the property class. 中文初译: 如果您只是使用驱动器作为财产,作为声明中的谓词,那么隐含地可以推断出,在三重中处于上游位置的东西属于财产类别。 00:15:30.250 EN: But sometimes you want to define a vocabulary, a schema, but without concrete data. And then it's good to be able to define properties also independently of concrete data. 中文初译: 但有时你想要定义一个词汇,一个计划,但是没有具体的数据。然后,很好的做法是能够定义属性, 与具体的数据无关。 00:15:44.440 EN: We can create hierarchies of properties. You should have discussed this also last week. So we can say that drives is a sub-property of controls. 中文初译: 我们可以建立属性等级。 您上周也应该讨论这个问题。 所以我们可以说驱动器是控制的一种次财产 。 00:15:56.080 EN: And of course it's an idealized model of the real world. Yeah, not always driving means controlling. There are some exemptions which you typically result in accidents, but in an idealized world, we assume that someone who drives a car also controls the car, right? 中文初译: 当然,这是现实世界的理想模式。对,开车并不总是意味着控制。有些例外通常会导致事故,但是在一个理想化的世界中,我们假设驾驶汽车的人也控制着汽车,对吗? 00:16:17.020 EN: And when we have then the statement max drives my blue Volkswagen Golf, we can infer that max also controls my blue Volkswagen Golf. 中文初译: 当我们有了最高声明 驱动我的蓝色大众高尔夫, 我们可以推断最大也控制了我的蓝色大众高尔夫。 00:16:28.560 EN: So this, what we have, this inferencing, subclass inferencing also then applies to sub-properties. We can also do inferencing over properties. 中文初译: 因此,我们所拥有的,这种推论,子类推论 也适用于次财产。我们也能够对属性进行推论。 00:16:45.040 EN: We can define the domain and range of properties. So, and that is a very important tool to restrict the number of reasonable triples. 中文初译: 我们可以定义属性的域和范围。因此,这是一个非常重要的工具,可以限制合理三重数的数量。 00:17:02.000 EN: So we, for example, the subject basically of a triple of a statement is the domain of a property and the object belongs to the range of the property. 中文初译: 例如,我们,比如说,一个声明的3倍的主题 基本上是财产的领域, 而物品属于该财产的范围。 00:17:17.280 EN: So here in that case, max belongs to the domain of drives and my blue Volkswagen Golf belongs to the range of the property drives. 中文初译: 因此,在这种情况下,最大属于驱动器的域, 我的蓝色大众高尔夫球属于财产驱动器的范围。 00:17:27.420 EN: And now we can explicitly state that for a certain property, we only want to allow instances of certain classes to be in the domain or in the range. 中文初译: 现在我们可以明确指出,对于某种财产, 我们只想要允许某些类别的情况 出现在域内或范围之内。 00:17:41.720 EN: So for the property drives, we can say that the domain is a person. That means everybody who drives something must be a person. 中文初译: 因此,对于财产驱动器,我们可以说域是个人。这意味着每个驾驶者必须是个人。 00:17:51.440 EN: And we can define the range. And we say here that everybody who drives or is driven by a person belongs to the class vehicle. 中文初译: 我们可以定义射程。我们在这里说, 驾驶或驾驶的每一个人 都属于类车辆。 00:18:04.500 EN: So we basically define what the subjects of triples must belong to, to which class and what the objects of triples with a certain property must belong to. 中文初译: 因此,我们基本上定义了什么是三重体必须属于的主体, 哪一类和什么是具有某种财产的三重体物体必须属于的。 00:18:16.100 EN: In that case here for drives, subject must be persons, objects must be vehicles. So similarly, we can also define the range to be a data type and that's what you see here. 中文初译: 在这种情况下,驱动器的主体必须是人,物体必须是车辆。同样,我们也可以将范围定义为数据类型,这就是你在这里看到的。 00:18:39.400 EN: H has a range non-negative integer. So that means that the H property should have values from the XML schema data type non-negative integer, which makes a lot of sense that you cannot add strings here or negative values. 中文初译: H 具有范围非负整数。 这意味着 H 属性应该包含来自“ XML” schema 数据类型非负整数的数值, 这非常说明您不能在此添加字符串或负整数。 00:18:59.200 EN: And in order to indicate the age of a person. It's important to understand that must have is not a constraint in the sense that my blue Volkswagen Golf is not a vehicle. 中文初译: 为了说明一个人的年龄。重要的是要明白,这必须不是限制,因为我的蓝色大众高尔夫不是车辆。 00:19:20.780 EN: Then the RDF statement above is illegal. It means you can still make any statement, but the inference would be that the subject belongs to a class. 中文初译: 那么上面的“ RDF” 声明是非法的。 这意味着你仍然可以做任何声明, 但推断是这个对象属于一个阶级。 00:19:33.220 EN: For example, it means that given X my blue Volkswagen Golf is used with X drives. We know that it is in vehicle in addition to whatever else it may be. 中文初译: 例如,它意味着,鉴于 X , 我的蓝色大众汽车高尔夫用于使用 X 驱动器。 我们知道它是在车辆上, 除了其他任何可能的东西之外。 00:19:46.760 EN: So it actually doesn't constrain you and that's a bit different maybe from databases. If you define a column in the database to be a foreign key relating to another database table, it's maybe a bit similar to the range. 中文初译: 所以它实际上并不限制你,这也许与数据库有点不同。如果你定义数据库中的一列是另一个数据库表格的外国密钥, 它可能和范围有点相似。 00:20:06.580 EN: Then you can only add values referring to this other table. Here you can still add any resource, but you should be aware that this resource will then become a person or a vehicle for example, by inference. 中文初译: 然后,您只能添加指向此表格的值。在这里您还可以添加任何资源,但您应该知道,该资源随后将成为一个人或工具,例如通过推论。 00:20:37.300 EN: Okay, and that's where the last slide which Markus already presented last week. So you can also add multiple range statements. 中文初译: 好的,这就是Markus上周已经展示的最后一张幻灯片。所以您也可以添加多个区域语句 。 00:20:48.960 EN: You can say that drives has a range car and you can also add drives has a range ship, but that actually means the range means that the instance of drives must belong to both classes, car and ship. 中文初译: 你可以说驱动器有一个射程汽车,也可以加上驱动器有一个射程船舶,但这意味着范围意味着驱动器必须属于两种类型,即汽车和船舶。 00:21:11.760 EN: And of course this doesn't make sense here. So that's why you would have to define that actually differently. So typically assigning several ranges to a property doesn't make sense, it's possible, but you should be aware that then the intersection actually of these two ranges is applied. 中文初译: 当然,这在这里是没有道理的。所以这就是为什么你必须用不同的定义来定义它的原因。 所以通常地说,给一个财产分配几个范围是不合理的, 这是可能的, 但你应该知道这两个范围的交叉点实际上被应用了。 00:21:37.760 EN: So the intersection between car and ship is then actually the range, it's not the union. And that's very important. 中文初译: 因此,汽车和船舶之间的交叉点实际上就是射程, 它不是联合体。这是非常重要的。 00:21:45.440 EN: So if you want to express that the object of the property has to be a car or a ship, a better expression would then be to define car as a subclass of vehicle, to define ship as a subclass of vehicle and then to apply the range of two drives, the range vehicle, basically to define the range for drives as vehicle. 中文初译: 所以,如果你想表达财产的客体必须是一辆车或一艘船,那么一个更好的表述是把汽车定义为车辆的子类,将船舶定义为车辆的子类,然后将船舶界定为车辆的子类,然后适用两个驱动器的范围,即射程车辆,基本上将驱动器的范围定义为车辆。 00:22:12.540 EN: And then it applies of course also to the subclasses. So that's a bit recapitulating what you discussed last week, as Markus told me. 中文初译: 然后它当然也适用于子类。所以这是你上星期讨论的 重述一下,正如马库斯告诉我的那样。 00:22:24.700 EN: And now we start continuing with some new information for today. We have also implicit knowledge. Once we define the domain and range of properties, we have to take care that using such properties could lead to unintended consequences. 中文初译: 现在我们开始为今天继续提供一些新信息。我们也有隐含的知识。一旦我们定义了属性的域和范围,我们必须小心使用这些属性可能导致意外后果。 00:22:42.720 EN: I already teased that a bit. So for example, if you have the schema, Mary 2 has the domain person and Mary 2 has the domain and the range person and Institute AKSW, for example, it's an Institute has the type institution. 中文初译: 我已经取笑过这一点了。比如说,如果你有这个计划, Mary 2 拥有域名, 而Mary 2 拥有域名, 靶员和AKSW研究所, 比如说,这是一个研究所,拥有型号机构。 00:23:04.500 EN: We are of course allowed to add a triple, Max is married to Institute AKSW. So RDF schema doesn't prohibit any triples, but RDF schema defines what the consequences of adding such weird triples would be. 中文初译: 我们当然可以增加三倍, Max是AKSW研究所的会员。所以“RDF”schema并不禁止任何三倍,但“RDF”schema定义了添加这种奇怪的三倍的后果。 00:23:23.660 EN: And the logical consequence would be that Institute AKSW is a person, belongs to the class person. And that of course would not make sense. 中文初译: 其逻辑后果是AKSW研究所是一个人,属于阶级。这当然不合理。 00:23:35.980 EN: So there are reasoners which actually infer such implicit knowledge. You can do it yourself, but reasoners or inference engines, they do this automatically, systematically and then they basically tell you such things and then you can discover, oh, something probably doesn't make sense in my knowledge base because I come to consequences which were unintended. 中文初译: 所以有些理性者可以实际推断出这种隐含的知识。你可以自己做, 但是理性者或推论者引擎, 他们这样做是自动的,系统化的, 然后他们基本上告诉你这些事情 然后你就可以发现, 哦,在我的知识库里有些东西可能没有意义, 因为我遇到了意外的后果。 00:24:02.260 EN: So this logical consequence here, Institute AKSW is a person would not make sense and then you would investigate what could be the reason for that. 中文初译: 因此,这个逻辑结果在这里,AKSW研究所是一个人 是不合理的, 然后你会调查什么可能是原因。 00:24:12.060 EN: And the reason for that is that Max is said to be married to Institute AKSW. So then you would have to remove this triple. 中文初译: 原因就是Max据说嫁给了AKSW研究所。所以你必须取消这三重任务。 00:24:27.620 EN: So there might also be false conclusions. Some people might be confused about a combination of specifying domain and range of a property and the hierarchy of classes. 中文初译: 因此,也可能存在错误的结论。 有些人可能会被混淆为将指定领域、财产范围以及等级分级混在一起。 00:24:39.120 EN: So let's look at an example. We have drives has as a range car and car is a subclass of a vehicle. So these two triples do not entail the following statement drives has a range vehicle. 中文初译: 让我们来举一个例子。我们有驱动器,像一个射程车一样, 汽车是一辆汽车的子类。所以这两辆三重车并不意味着下面的对数驱动器有一辆射程车。 00:24:58.080 EN: That's very important. So it works the other way around. If drives would be applied to vehicle, then of course you can also apply to cars, but not the other way around, right? So not to super classes. 中文初译: 这很重要。 所以它反之而行。 如果驱动器会应用到车辆上, 那么当然也可以应用到汽车上, 而不是另一边,对吧?所以不要应用到超级类。 00:25:19.120 EN: The range is always applied to the more specific class here in that case, car, and that then it's not applied to vehicle. 中文初译: 范围总是适用于这里更具体的类别, 在这样的情况下,汽车, 然后它不适用于车辆。 00:25:33.720 EN: So we don't gain new terminological knowledge. Still of any concrete triple having the predicate drives, we know that object is of type vehicle. Since we know that every object of a triple having drives in the predicate position is a car, then since car is a subclass of vehicle, it also must be a vehicle, right? 中文初译: 所以我们没有获得新的术语学知识。 即便有任何混凝土的三联驱动器, 我们也知道该物体是型号车辆。 既然我们知道在上游位置上,每个有三连驱动器的物体都是一辆车, 那么既然汽车是一个小类车辆, 它也必须是一辆车,对吧? 00:26:02.790 EN: So that's important. Then coming to container classes as the next topic. And they are a bit debated. They are defined in the standard, but some people or some discourage its use a bit. 中文初译: 因此,这很重要。 然后来到集装箱类作为下一个主题。 他们有些争论了一下。 在标准中定义了它们, 但有些人或一些人不鼓励使用它。 00:26:33.710 EN: So we have this class RDFS container, which is a superclass for the containers defined by RDF, and the containers are sequence, bag, and alternative. 中文初译: 因此,我们有了这个“RDFS”类集装箱,这是“RDF”定义的集装箱的超级级, 集装箱是序列、袋和替代物。 00:26:44.340 EN: In short, seek, bag, and alt. And let's have a look. Also, there is a property defined to work with these containers. 中文初译: 简而言之,找寻、袋和替代物,让我们来看看。还有一个财产 被确定与这些容器合作。 00:26:58.940 EN: There is the first class RDFS container membership property defined as a class, which contains all properties that are used to state that a resource is a member of a container. 中文初译: 第一类“RDFS”集装箱成员财产被界定为一种类别,包含用于说明资源是集装箱成员的所有财产。 00:27:13.040 EN: And they have these underscore one, underscore two, and so on names, yeah? So these are a sequence of properties, which is indefinite. 中文初译: 他们有这些强调1,强调2,等等 在名称上,是的。所以这些是一系列的属性, 是无限期的。 00:27:27.600 EN: So you can have many of these underscore properties with a numeric value in order to indicate that something belongs to a container. 中文初译: 因此,您可以拥有许多这些带有数字值的加分属性,以表明某物属于容器。 00:27:41.220 EN: So then we have the RDFS member property, which is a super property for all properties of a container, for all container membership properties. 中文初译: 因此,我们拥有“RDFS”成员财产,这是集装箱所有财产的超级财产,也是所有集装箱成员财产的超级财产。 00:27:54.080 EN: So every instance of container membership property is a sub-property of the RDFS member property. But let's maybe look at some examples to illustrate how that works. 中文初译: 因此,每个关于集装箱成员财产的例子都是“RDFS”成员财产的次级财产。但是我们不妨看看一些例子来说明它是如何运作的。 00:28:18.540 EN: Actually, the examples come a bit later or are missing here. Yeah, we come back to these examples for the containers later on. 中文初译: 事实上,这些例子迟些时才会出现或在这里丢失。 00:28:36.120 EN: Another feature of RDFS schema, RDFS, RDF schema, is refacation. So, and refacation is used to state that, to make statements about statements. 中文初译: “RDFS”schema、“RDFS”schema的另一个特征是重构。因此,用“RDF”schema 来表述发言内容。 00:28:54.080 EN: For example, we have a sentence, the detective supposes that the butler killed the gardener, right? So how can we represent this information in RDF? 中文初译: 比如说,我们有一个句子 侦探假设管家杀了园丁 对吧? 00:29:05.060 EN: So we could, for example, use the first triple here, detective supposes, and then at the string, the butler killed the gardener. 中文初译: 比如说,我们可以使用第一个三重 侦探假设 然后在弦上 管家杀了园丁 00:29:13.920 EN: So this would be possible to encode the information as a string, but we cannot really search then for statements, like suspicions of the detective, for example. 中文初译: 这样就可以把信息编码为字符串, 但我们不能真正去寻找证词, 比如侦探的怀疑。 00:29:28.700 EN: Another possibility would be to create a new identifier, a new resource identifier, the butler killed the gardener for a resource, and say that the detective supposes this new resource, but both these ways are unsatisfactory. 中文初译: 另一种可能性是建立一个新的识别资料,一个新的资源识别资料, 管家为了一个资源杀了园丁, 并且说侦探认为这个新资源, 但这两种方式都令人不满意。 00:29:50.140 EN: What we actually want to do is, we want to talk about the triple butler property killed and object gardener, right? 中文初译: 我们其实想做的是, 我们想要谈谈 三层管家财产被杀 和物件园丁,对不对? 00:29:58.600 EN: So we want to talk about this property, this triple, the statement itself. How can we do that? So, and for that purpose, we can use the class RDFS statement to refine statements. 中文初译: 所以我们想谈谈这个属性, 这个三重, 声明本身。 我们如何做呢? 所以,为了这个目的, 我们可以使用“ RDFS” 类声明来完善语句。 00:30:20.990 EN: And the problem maybe with the statement, butler killed the gardener, we could of course add this triple to our knowledge base, but we actually don't know if it's true. 中文初译: 也许声明的问题, 管家杀了园丁, 我们当然可以增加这三倍的知识库, 但我们实际上不知道这是不是真的。 00:30:32.190 EN: So we should only add information to the knowledge base, which we know that it's a fact, that it's true. If it's just a hypothesis, then maybe it should not be directly added to the knowledge base. 中文初译: 因此,我们只应该在知识库中添加信息, 我们知道这是事实, 这是真实的。如果这只是一个假设, 那么也许它不应该直接加入到知识库中。 00:30:47.070 EN: So, and there is the possibility to define RDF statements in RDF itself using the properties RDF subject, which defines an RDFS resource, which is the subject of the statement, then RDF predicate, which defines a property, which is the predicate of the statement, and RDF object, which defines a resource, which is the object of the statement. 中文初译: 因此,可以使用“RDF”本身的属性“RDF”主体来定义“RDF”中的“RDF语句,该元素界定了“RDF”资源,这是该语句的主题,然后是“RDF”上游,它界定了一种财产,即该语句的谓词,以及“RDF”对象,它界定了一种资源,即声明的对象。 00:31:15.650 EN: And using refications, we can see an RDF triple as a resource and formulate facts about it. For example, a theory that has not been proven yet. 中文初译: 使用重新解释,我们可以看到“RDF”的三倍资源,并编造关于它的事实。例如,这个理论还没有被证实。 00:31:26.550 EN: So let's look how this works. So here we say, we see exactly our example. We have detective, supposes, and then we have a theory. 中文初译: 让我们来看看这是怎么回事。我们在这里说, 我们可以看到我们的榜样。 我们有侦探,假设,然后我们有一个理论。 00:31:39.500 EN: You see here, this underscore means it's a blank note. So the theory itself doesn't have a worldwide unique identifier, but we could also create one here. In that case, it's a blank note. 中文初译: 你看,这里,这个强调的意思是它是一个空白的音符。所以理论本身没有全世界独特的识别符号, 但我们也可以在这里创建一个。如果是这样的话, 它是一个空白的音符。 00:31:53.850 EN: Then we say that this theory is of type RDFS, RDF statement, yeah? And then we indicate the subject of the theory statement, the predicate of the theory statement, and the object of the theory statement. 中文初译: 然后我们说这个理论是“RDFS”类型,“RDF”声明。然后我们指出理论声明的主题,理论声明的谓词,以及理论声明的目标。 00:32:09.350 EN: Subject is butler, predicate is has killed, and object is gardener, yeah? And then we can, in addition, add more metadata, more information here. 中文初译: 主体是管家,上游已经死亡,对象就是园丁,对吗?然后我们可以在这里添加更多的元数据, 更多的信息。 00:32:20.850 EN: In that case, we say we add a state, and we indicate that it's unproofed as a literal. So this particular theory is unproofed, yeah? 中文初译: 这样的话,我们就说我们增加一个状态, 我们表示它没有被一字不差地标注。所以这个理论是没有被校正的,是吗? 00:32:33.290 EN: And for example, if you would create a criminalistic knowledge base, you can then distinguish proven theories from unproven ones. 中文初译: 例如,如果你能创建一个犯罪学知识库, 你可以把被证明的理论和未经证实的理论区分开来。 00:32:44.880 EN: So note that the following statement is not a logical consequence of this. So butler killed the gardener is not in our knowledge base itself, yeah? 中文初译: 请注意,以下声明不是这个现象的逻辑后果。因此,巴特勒杀死园丁并不是我们的知识基础本身,是么? 00:32:54.960 EN: We didn't add the triple itself to the knowledge base. We just added a description of the statement to the knowledge base, and that's exactly this part here. 中文初译: 我们没有在知识库中增加三分之一。我们刚刚在知识库中增加了对声明的描述, 这正是这部分。 00:33:05.540 EN: And that allows us to make statements about statements, and that's called reification. It's a bit, of course, complicated, yeah? 中文初译: 这让我们可以做关于发言的演讲, 这就是所谓的重新解释。这当然有点复杂,是么? 00:33:17.360 EN: Because for one statement, for one triple, we basically have to add four additional triples to the knowledge base. 中文初译: 因为对于一个说法,对于一个三者, 我们基本上必须增加四个额外的三者 知识基础。 00:33:24.660 EN: So the space, the storage requirements explode a bit if you have a lot of these statements in our knowledge base. 中文初译: 所以空间,存储要求 有点爆炸 如果你在我们的知识库里有很多这些声明的话 00:33:38.170 EN: Let's look how this looks as a graph. So here we have the graph for this small knowledge base. So we have the blank note in the middle for our statement. 中文初译: 让我们看看这个图象如何。 因此我们这里有这个小知识库的图表。 所以我们中间有一个空白的纸条用于我们的语句 。 00:33:50.310 EN: It belongs to the class RDFS statement. It has the state unproven here, the literal. The detective supposes this theory, and it has as a subject butler, as a predicate killed, and as an object gardener, yeah? 中文初译: 它属于“RDFS”类声明。它有国家在这里没有证明, 字面上。侦探认为这个理论, 它作为一个主体管家, 被杀死的谓词, 和作为花匠,是吗? 00:34:07.890 EN: So that's the visual notation for this reification. Yeah, another feature of RDFS is adding supplementary information. 中文初译: 这就是这个重新校正的视觉标记。 是的,“RDFS”的另一个特征是添加补充信息。 00:34:22.780 EN: So RDF schema gives us the possibility to add additional information to resources using the following properties. 中文初译: 因此,“RDF”方案使我们有可能利用下列特性在资源中添加补充信息。 00:34:32.800 EN: RDFS label, which can be used to give a human readable name for a resource. You can also attach labels in different languages. 中文初译: “ RDFS” 标签,可用于给资源提供一个人类可读的名称。您也可以用不同语言附加标签。 00:34:41.420 EN: For example, you can have English, German, Russian, and French, Spanish, and so on labels in different languages, and that allows to have multilingual vocabularies, RDF schema, or ontologies. 中文初译: 例如,您可以使用英文、德文、俄文、法文、西班牙文等不同语言的标签,并且可以使用多语种词汇、“RDF”计划或本体论。 00:35:02.020 EN: Labels should be short names, yeah, for a resource. For longer information, we have RDFS comment. So we can add longer comments, definitions, longer explanations, basically, to resources. 中文初译: 标签应该是短名, 是的, 对于资源来说。 对于更长的信息, 我们有“ RDFS ” 注释。 这样我们可以在资源中添加更长的评论、 定义、 更长的解释 。 00:35:19.760 EN: Another property which is defined in RDF schema is C also. This points to URIs where additional information about a resource can be found. 中文初译: “RDF” schema中定义的另一个属性也是C。这指向“URIs”,可以找到关于资源的额外信息 。 00:35:29.560 EN: So this can be, for example, webpages or other vocabularies or ontologies. And the last one here is is defined by, which points to an URI where the resource is defined. 中文初译: 举例来说,这可以是网页或其他词汇或本体。这里最后一个是由它定义的, 它指向资源定义的“ URI ” 。 00:35:45.050 EN: And it's actually a sub-property of RDFSC also, yeah? Of course, if something is defined somewhere, it's also a reference where to look for more information. 中文初译: 它实际上也是皇家国防和安全部队的次级财产,是的。当然,如果某处有定义的话,它也是一种参考, 用来查找更多信息。 00:35:57.930 EN: So and here we have an example for Volkswagen Golf. We could attach the label, Fauvee Golf, yeah? And you could attach labels in, for example, also Cyrillic or Arabic if you have different transliterations of the name. 中文初译: 这里我们有一个大众高尔夫的范例。 我们可以附加标签, Fauvee Golf, 是的吗? 并且您也可以在西里尔语或阿拉伯语中加上标签, 如果你对名称有不同的转音。 00:36:18.030 EN: Then we can add a comment here and then write a small text. We can then add a C also link, for example, to the Wikipedia page about Volkswagen Golf. 中文初译: 然后我们在这里添加一个注释, 然后写一个小文本。 然后我们可以在维基百科页面上加入 C 链接, 例如关于大众高尔夫的链接 。 00:36:32.880 EN: And we can add is defined by and maybe there is a data set published by Volkswagen itself. And then we can refer to this data set and where you can find maybe the product specification and so on. 中文初译: 我们可以添加由大众传媒自己 所定义的数据集。 然后我们可以引用这个数据集, 以及你能找到产品规格等等的地方。 00:36:58.300 EN: So the advantage of using these properties is that additional information is represented as structured RDF2. So the metadata is encoded directly in RDF as well. 中文初译: 因此,使用这些属性的优点是 额外信息被表述为结构化 RDF2。 所以元数据也直接编码在“RDF”中。 00:37:12.440 EN: And that's quite beneficial. So you don't need to add comments. For example, in XML, there are these comments or also HTML comment sections. 中文初译: 这非常有益。所以你不需要添加评论。例如,在“XML”中,有这些评论或HTML评论部分。 00:37:23.140 EN: So this is not required. You can directly add the supplementary information in RDF itself. So you don't have to use a different formalism for it. 中文初译: 因此不需要这样做。 您可以直接在“ RDF” 本身添加补充信息。 这样您不必使用不同的形式主义 。 00:37:37.040 EN: And now let's look at a bit, a larger example of a small knowledge base. We want to implement a system which should be able to calculate the needed amount of an asset or base to neutralize a given solution. 中文初译: 现在让我们来看一下,一个小知识库的更大例子。我们想实施一个系统, 这个系统应该能够计算一个资产或基地所需的数量, 来抵消给定的解决方案。 00:37:53.800 EN: That's an example from chemistry. Therefore, our system has to handle information about assets and basis. Maybe you all remember from school what asset and basis are. 中文初译: 这是化学的一个例子。因此,我们的系统必须处理资产和基础的信息。也许你们都记得学校里的资产和基础是什么。 00:38:06.180 EN: For this information, we define our own schema for our data. Yeah, a domain-specific schema. Now RDF, with RDF schema, we can define domain-specific schema term. 中文初译: 对于这个信息,我们定义了我们的数据模式。是的,一个特定域的模式。现在的“RDF”,用“RDF”的模式,我们可以定义特定域的模式术语。 00:38:19.020 EN: So we start by defining asset and base as own classes. So that's what we have here. Asset is of type RDFS class. Base is of type RDFS class. 中文初译: 因此,我们首先将资产和基地定义为属于自己的类别。这就是我们这里的情况。资产是“RDFS”类。基础是“RDFS”类。 00:38:31.200 EN: Both can be described as chemical compounds. So we define this as an own class and the two other subclasses of it. 中文初译: 两者都可称为化学化合物。所以我们把它定义为一种自己的类别和另外两种子类别。 00:38:39.840 EN: So we have the class now chemical compound is also a class. And we define asset as a subclass of chemical compound and base as a subclass of chemical compound. 中文初译: 所以我们现在有了化学化合物类, 也是一种类别。我们把资产定义为化学化合物的子类, 把基地定义为化学化合物的子类。 00:38:57.730 EN: Then we can define some instances. So for example, we define hydrogen chloride, HCl. Which is the short chemical name for it. 中文初译: 然后我们可以定义一些例子。例如,我们定义氯化氢,HCl。这是它的简短化学名称。 00:39:14.690 EN: We define phosphoric acid, which is H3PO4. We define sodium hydroxide, which is NaOH. And we define calcium hydroxide, which has the short chemical name CaOH2. 中文初译: 我们定义磷酸,即H3PO4。我们定义氢氧化钠,即NaOH。我们定义氢氧化钙,即CaOH2,其化学名称简称为CaOH2。 00:39:34.610 EN: And we use these chemical short names for the molecules. As identifiers, resource identifiers here. In our own namespace, which has here the prefix X, like example. 中文初译: 我们用这些化学短名来识别分子。作为识别符号和资源标识符。 在我们自己的命名空间里, 这里有前缀 X, 例如 。 00:39:50.250 EN: So we define them, all the first two hydrogen chloride phosphoric acid as acids and sodium hydroxide and calcium hydroxide as bases. 中文初译: 因此我们定义了它们,所有前两个氯化磷酸氢作为酸,氢氧化钠和氢氧化钙作为碱基。 00:40:05.610 EN: Yeah, so two acids and two bases here in our example. Which we can then add to our knowledge base. So we now have the following small knowledge base. 中文初译: 是的,我们的例子中有两个酸和两个基础。然后我们可以增加我们的知识库。所以我们现在有以下的小型知识库。 00:40:18.700 EN: So we have acid subclass of chemical compound, base subclass of chemical compound. We have hydrogen chloride as an acid. 中文初译: 我们有酸化化学化合物的亚类,基级化学化合物的子类。我们有氯化氢作为一种酸。 00:40:33.740 EN: And we have phosphoric acid as an acid and sodium hydroxide and calcium hydroxide. So that's how our knowledge base looks like. 中文初译: 我们有磷酸作为一种酸,氢氧化钠和氢氧化钙。所以我们的知识基础就是这样的。 00:41:02.230 EN: So now we can define the molar mass, which is an important property in chemistry. After creating the classes, we want to add a typical chemical property like molar mass. 中文初译: 因此我们现在可以定义摩尔质量,这是化学中的重要属性。在创建这些类别后,我们想添加一个典型的化学特性,比如摩尔质量。 00:41:15.530 EN: We first define it as a property. So has molar mass is of type RDFS property. And then every chemical compound can have an information about its molar mass. 中文初译: 我们首先将它定义为一种属性。 摩尔质量属于“ RDFS ” 类型属性。 然后每一个化学化合物都可以获得关于它的摩尔质量的信息 。 00:41:32.930 EN: So has molar mass has a domain chemical compound. So that we can apply it to chemical compounds in particular to acids and bases because they were defined as subclasses of chemical compound. 中文初译: 摩尔质量也有一个域化学化合物。这样我们才能将其应用于化学化合物,特别是酸和碱,因为它们被定义为化学化合物的亚类。 00:41:47.910 EN: The molar mass itself could be defined as an own data type with some additional information. But for our small example, we want to use a simple existing floating point data type. 中文初译: 摩尔质量本身可以被定义为带有一些补充信息的数据类型。 但是,对于我们的小例子来说,我们想要使用简单的现有浮点数据类型。 00:41:58.670 EN: So the range of molar mass of the X has molar mass property is float. And we use float from the XML schema data type. 中文初译: 所以 X 的摩尔质量范围有摩尔质量属性。 我们使用“ XML” schema 数据类型的浮点 。 00:42:08.350 EN: So has molar mass has the range XSD float, which is a floating point number. And then we can define our sodium hydroxide to have the molar mass of 39.9971 by indicating it in that way here. 中文初译: 摩尔质量的 XSD 范围是浮点数,而摩尔质量是XSD 浮点数。然后我们可以定义我们的氢氧化钠,将摩尔质量定为39.9971, 以此表示它在这里。 00:42:42.280 EN: As a general difference between acid and base, we want to define that an acid can donate protons and the base can accept protons. 中文初译: 作为酸和碱之间的一般差异,我们想确定一种酸可以捐赠质子,而碱可以接受质子。 00:42:51.100 EN: That's the difference between acid and base and the definition of acid and base. Therefore, we want to define two properties with which we can store how many protons can be accepted or donated. 中文初译: 这就是酸与碱的区别, 以及酸与碱的定义。 因此, 我们要定义两个属性, 我们可以储存多少质子可以被接受或捐赠。 00:43:05.440 EN: So that's why we define the property can donate protons. First, we apply can donate protons to acids. That's why it has the domain acid here and the range is integer. 中文初译: 所以我们定义属性可以捐赠质子的原因。 首先,我们应用这些属性可以捐赠质子到酸中。 这就是为什么它在这里有域酸, 范围是整数的。 00:43:21.940 EN: Actually it should be non-negative, it could be or should be even non-negative integer since we have here can donate protons. 中文初译: 实际上,它应该是非负性的, 可能是甚至应该是非负性的整数, 因为在这里我们可以捐赠质子。 00:43:32.960 EN: We could of course also model this differently. Yeah, so when you model things, there are always also different possibilities of modeling that. 中文初译: 当然,我们也可以用不同的模型来模拟。所以当你做模型的时候, 也总是有不同的可能性来模拟它。 00:43:42.020 EN: Here we define two properties, can donate protons and can accept protons. We could also do this with a single property. 中文初译: 在此我们定义两个属性, 可以捐赠质子, 也可以接受质子。 我们也可以使用单一的属性来做到这一点 。 00:43:50.300 EN: We could define a single property, can donate or accept protons and then the range would be integer and then when it is negative, then it donates, when it's positive, it accepts. 中文初译: 我们可以定义单一的属性, 可以捐赠或接受质子, 然后范围将是整数 当它是负数时, 然后它会捐献,当它是正数时, 它会接受。 00:44:08.260 EN: This would be another way how to model that. But this way here is maybe less ambiguous and easier. So then we have the second property, can accept protons. 中文初译: 这将是另一种模式。但这样,我们也许就比较模糊和容易了。所以我们有了第二个财产,可以接受质子。 00:44:20.100 EN: It applies to bases and it has the range also integer. The advantage of this modeling is that we have the two properties and the properties have the domain acid and base. 中文初译: 它适用于基座, 并且它的射程也具有整数。 这个模型的优点是, 我们拥有两个属性, 而特性有域酸和基底 。 00:44:35.020 EN: If we had only one property and we distinguished whether something is accepting or donating, then we would have to attach it to the chemical compound. 中文初译: 如果我们只有一个财产,并且区分是接受还是捐赠, 那么我们就不得不把它附在化学化合物上。 00:44:46.120 EN: So we would have to define the domain for it as chemical compound and distinguish whether it accepts or donates based on the positive or negativeness of the integer value. 中文初译: 因此,我们必须将其域名定义为化学化合物,并根据整数值的正负值来区分它接受还是捐赠。 00:45:04.060 EN: Here we modeled it with two properties. Then we indicate the acids and bases and we indicate the strength of them. 中文初译: 在这里,我们用两个属性来模拟它。然后我们指出酸和碱,并表明它们的强度。 00:45:18.120 EN: So basically meaning the number of protons they can donate or accept. So for hydrogen chloride, it can donate one proton. 中文初译: 所以基本上是指他们可以捐赠或接受的质子数量。对于氯化氢来说,它可以捐赠一个质子。 00:45:29.820 EN: So we indicated here one. For spheric acids can donate three protons. Then we have sodium hydroxide, it can accept one proton and calcium hydroxide can accept two protons. 中文初译: 我们在这里指出一个。因为球酸可以捐赠三个质子。然后我们有了氢氧化钠,它可以接受一个质子,氢氧化钙可以接受两个质子。 00:45:48.280 EN: So we added this to the knowledge base. Now we can define a solution. Our program has to work with solutions which contain different amounts of chemical compounds. 中文初译: 因此,我们把这个添加到知识库中。 现在我们可以定义一个解决方案。我们的方案必须使用含有不同数量化学化合物的解决方案。 00:46:00.130 EN: So we have to define the class solution and the property describing its mass. And that's what you see here. So we have a solution defined as a also belong to the RDFS class. 中文初译: 所以我们必须定义阶级解决方案 和描述质量的属性。 这就是你在这里看到的。所以我们有一个也属于“RDFS”类的解决方案。 00:46:15.460 EN: We have then the property has mass defined as a property. We have domain, it's applied to solution and a range is a floating point number here again. 中文初译: 然后,我们有了财产质量被定义为财产。我们有域,它被用于解决问题,一个范围又是一个浮动点编号。 00:46:32.710 EN: Now we define ingredients. Our program has to work with solutions which contain different amounts of chemical compounds. 中文初译: 现在我们来定义成分。我们的计划必须 与含有不同数量化学化合物的解决方案合作。 00:46:44.040 EN: However, a single property which describes that the chemical compound is inside a solution is not expressive enough because we need to also the possibility to give information about the amount of this compound. 中文初译: 然而,一个说明化学化合物在一种溶液中的单一财产不够明确,因为我们还需要有可能提供关于该化合物数量的信息。 00:46:58.220 EN: So we have to define the class ingredient and the property for describing the amount in percentages. Well, that's what you see here. 中文初译: 所以我们必须定义等级成分和财产 来用百分比来描述数量。这就是你在这里看到的。 00:47:08.340 EN: We define ingredient as a class and then has amount as a property and has amount has the domain ingredient and the range again, a floating point number which is then in gram. 中文初译: 我们把成分定义为一个类别,然后将数量作为一个属性,而数量则具有域成分和范围,即浮动点数,然后以克表示。 00:47:30.080 EN: Additional, this class needs two properties to be connected to a solution and a chemical compound. So, and these two properties are is part of and contains. 中文初译: 此外,这一类需要两个特性才能连接到溶液和一个化学化合物。所以,这两个特性是其中的一部分并包含在内。 00:47:43.370 EN: So is part of as a property, it has the domain ingredient and the range solution and contains is also a property, has a domain ingredient and the range chemical compound. 中文初译: 同样,作为财产的一部分,它具有域成份和射程溶液,包含也是一项属性、有域成份和射程化学化合物。 00:48:01.910 EN: Now we are able to express with our schema all the information which is needed to solve a typical task in chemistry and the typical task could be you have 100 gram of a solution which consists of 20% phosphoric acid. 中文初译: 现在我们可以用我们的计划来表达 解决典型化学任务所需要的所有信息 而典型的任务可能是 你拥有100克的溶液 含20%的磷酸 00:48:20.250 EN: How much of a 10% sodium hydroxide solution do you need to neutralize this solution? So we add then the information into our knowledge base which uses the schema which we have defined now. 中文初译: 需要多少10%氢氧化钠溶液来抵消这一溶液? 因此,我们将信息添加到我们的知识库中,它使用我们现在定义的模式。 00:48:43.190 EN: So we define given solution as a solution. It has a mass of 100 grams here. Of course, to do this property properly, we should have indicated also a data type gram here. 中文初译: 因此,我们定义给定的解决方案是一个解决方案。 它在这里有100克的质量。 当然, 要正确处理这个属性, 我们也应该在此显示一个数据类型 。 00:49:00.510 EN: We simplified that here a bit in this example. Then we define our phosphoric acid ingredient, 20% ingredient as an ingredient. It has the amount 20% here and it contains phosphoric acid and is part of our given solution which we defined before. 中文初译: 我们在此稍作简化。 然后我们定义磷酸成分, 20%的成分为一种成分。 它的含量是20%, 它含有磷酸, 是我们以前定义的溶液的一部分。 00:49:32.590 EN: Now we define search solution and this belongs to the class solution and we define our sodium hydroxide ingredient 10% which has an amount of 10% of sodium hydroxide and is part of the search solution. 中文初译: 现在,我们定义搜索解决方案,这属于类解决方案。 我们定义了氢氧化钠成分的10%。 氢氧化钠含量为10%,是搜索解决方案的一部分。 00:50:06.070 EN: And with this information, now we could compute how much is required. So the program, our software program could now query the following needed information from our knowledge base. 中文初译: 有了这些信息,我们现在可以计算出需要多少。所以这个程序,我们的软件程序 现在可以从我们的知识库中查询以下所需的信息。 00:50:20.010 EN: So the molar mass is retrieved and how many protons can be donated by phosphoric acids and the same for sodium hydroxide, also the molar mass and how many protons can be accepted. 中文初译: 因此,取回了摩尔质量,有多少质子可以用磷酸捐赠,氢氧化钠的质子数量相同,还有摩尔质量和有多少质子可以被接受。 00:50:44.500 EN: And relying with this information and using our schema, the program can then compute the result for the given task. 中文初译: 依靠这些信息和我们的计划, 这个程序就可以计算给定任务的结果。 00:50:53.630 EN: And in that case, 244.9 gram of the sodium hydroxide solution are required for neutralizing the given phosphoric acid solution. 中文初译: 在这种情况下,需要244.9克氢氧化钠溶液来消除特定磷酸溶液。 00:51:07.700 EN: So the computation has to be done still by a software program. Also retrieving the information from the knowledge base is something which we can do with the query language. 中文初译: 因此计算必须仍然由软件程序完成。 另外从知识库检索信息是我们可以用查询语言做的工作。 00:51:19.380 EN: So that will be one of the next lectures where we discuss the SPARQL query language which allows to query information. 中文初译: 因此,这将是下一个讲座之一, 我们在这里讨论“SPARQL”查询语言, 以便查询信息。 00:51:31.160 EN: And that was a small example illustrating basically how RDF schema can be used to model a domain here in that case, chemistry, and then can be applied of course, maybe for such a simple example, it's a bit over coil of creating such a knowledge base, but imagine you have a library and knowledge base with all possible assets and basis. 中文初译: 这是一个小例子, 基本地说明“RDF”的图案 如何可以用来建模一个域, 在那样的情况下,化学,然后当然也可以应用, 也许对于这样一个简单的例子来说, 它比创造这样的知识库 还要多一点, 但是想象你有一个图书馆和知识库, 拥有所有可能的资产和基础。 00:51:58.560 EN: Then it makes actually sense to have such a knowledge base and to retrieve information from such a knowledge base. 中文初译: 然后,拥有这样一个知识库和从这样的知识库检索信息实际上是有道理的。 00:52:05.940 EN: You could of course also do this with a relational database, but especially when you want to add more information then such a knowledge base is more flexible. 中文初译: 当然,也可以使用关系数据库这样做,但特别是当你想添加更多信息时,那么这样的知识库就更加灵活。 00:52:18.460 EN: And we see some examples of that later on. So what we have learned is basically that RDF and RDF schema allow us to define two types of knowledge, assertional knowledge and terminological knowledge. 中文初译: 我们稍后会看到这方面的一些例子。所以我们所学到的基本上就是“RDF”和“RDF”计划让我们可以定义两种类型的知识,即主张知识和术语学知识。 00:52:35.710 EN: So RDF schema extends RDF with a special vocabulary for terminological knowledge. And in this picture here, you see the distinction between the two. 中文初译: 所以“RDF”schema 扩展了“RDF ” , 并配有术语学知识的特殊词汇。 在这张图片中, 你可以看到两者的区别。 00:52:46.390 EN: The assertional knowledge are basically facts from the real world, yeah? While the terminological knowledge allows to add some structure, some schema to it. 中文初译: 断言知识基本上是真实世界的事实,是么?虽然术语学知识允许增加某种结构,但有些计划。 00:52:59.010 EN: So you can see this also a bit similar to relational databases. You have the data inside the relations, the tables in your database. 中文初译: 因此,您可以看到它也和关系数据库有点相似。您在关系中拥有数据,在数据库中有表格。 00:53:08.030 EN: That's the assertional knowledge. And then you have the definitions of the tables and the columns, and that's the terminological or the schema knowledge. 中文初译: 这就是所谓的知识。然后,你有表格和柱子的定义, 那就是术语学或系统学的知识。 00:53:19.740 EN: RDF schema and also OWL are a bit more expressive and can represent knowledge. For example, inference is possible. 中文初译: “RDF”的图案和“OWL”更能表达,可以代表知识。例如,推论是可能的。 00:53:28.960 EN: That's also distinguish it from a relational data representation. But relational data can be also easily expressed in RDF and RDF schema. 中文初译: 这也将其与关系数据表示法区分开来。但关系数据也可以很容易地用“RDF”和“RDF”的模型表达。 00:53:39.220 EN: We will also see that later in another lecture how that can be done. What are the limitations of RDF schema? So RDF schema can be used as a lightweight language for defining a vocabulary, also called ontology used in RDF graphs. 中文初译: 我们也可以在另一场演讲中看到如何做到这一点。 “RDF”的“Schema”有什么局限性?因此,“RDF”的“schema”可以作为一种轻量级语言来定义词汇,也被称为“RDF”图表中使用的文理学。 00:54:00.920 EN: However, it has some limitations regarding the possibility of formulating ontologies. So there is missing expressivity. 中文初译: 然而,它对于制定本体论的可能性有一些限制,因此缺乏直观性。 00:54:11.040 EN: For example, it is not possible to define a negation of an expression. For example, it's not possible to state that the domain of a property does not contain a certain class. 中文初译: 例如,无法定义对表达式的否定。 例如, 无法声明财产领域不包含特定类别 。 00:54:24.480 EN: This negation is not possible. It's also not possible to define cardinalities. So for example, we cannot say that a person can only be married to one other person. 中文初译: 这种否定是不可能的, 也不可能定义基本因素。例如, 我们不能说一个人只能嫁给另一个人。 00:54:36.980 EN: This is not possible in RDF schema. We will learn in OWL, in the web ontology language, OWL also in one of the next lectures, how that can be done. 中文初译: 这在“RDF”中是不可能的。我们将用“OWL”来学习网络文学语言“OWL”和“OWL”在接下来的讲座中学习,如何做到这一点。 00:54:46.820 EN: So OWL actually adds more and solves than some of the lacking expressivity problems of RDF schema. It's also not possible to define a set of classes. 中文初译: 所以“OWL”实际上增加了更多和更多的解决方案, 而不是一些缺乏表达性的“RDF”计划问题。 也不可能定义一组类别。 00:54:59.260 EN: For example, we cannot express that the domain of a property should be one or another class. We have to create a new superclass of both, which can be used as a domain of the property. 中文初译: 例如,我们无法表示财产领域应该是一类或另一类。 我们必须建立一个两者皆有的新的超级类别,可以用作财产领域。 00:55:10.040 EN: I've shown you this with drives and vehicle. So we have applied drives to vehicle. We cannot say drives has the domain car and has the domain, what did we have, ship, because this would actually be the intersection of both. 中文初译: 我用驱动器和车辆向你们展示了这个。所以我们把驱动器应用到车辆上。我们不能说驱动器拥有域盘汽车, 并且拥有域盘, 我们拥有什么, 船舶, 因为事实上这是两者的交叉点。 00:55:29.660 EN: And there are not many cars which also ships. Some are, but not many. So for that, we also need more expressivity. 中文初译: 也没有多少汽车也是船只。有些是,但并不是很多。因此,我们也需要更多表达性。 00:55:40.860 EN: Also adding metadata of the schema, like versioning and so on, cannot yet be added. And that's something which is part of the web ontology language, which we discuss in a later lecture, sorry. 中文初译: 此外,还无法添加模型元数据,比如版本等等。这是网络本体语言的一部分,我们在稍后的演讲中讨论这个问题,抱歉。 00:56:03.380 EN: Yeah, so summary, RDF schema, short RDFS can be used to express terminological knowledge by defining classes and properties. 中文初译: 简称RDF 可以用来表达术语学知识 通过定义分类和属性 00:56:13.630 EN: The classes and properties can be arranged in hierarchies. The domain and range of properties can be defined. The schema allows the inference of knowledge, which has been defined implicitly. 中文初译: 类别和属性可以按等级排列。可以界定属性的域和范围。该体系允许对知识进行推断,而知识是隐含定义的。 00:56:27.480 EN: So we can, for example, discover that something is part of a class, which is actually the superclass of it, or also that use, exploit a relationship between properties, the sub-properties. 中文初译: 因此,举例来说,我们可以发现某物是某一类的一部分, 实际上是该类的超级阶级, 或者也是它使用的超级阶级, 利用属性与亚财产之间的关系。 00:56:44.620 EN: For example, being married to would also mean, for example, could be a sub-property of knowing someone. If you should be married to a person which you also know, that's why being married to could be a sub-property of knowing someone. 中文初译: 比如说,结婚也意味着,比如说,可能是一种认识某人的次级财产。 如果你应该嫁给一个你也知道的人,这就是为什么结婚可能是认识某人的次级财产。 00:57:05.140 EN: And then being married to would implicitly also mean if it's defined as a sub-property of knowing that you know the person you are married to. 中文初译: 然后结婚 隐含地意味着 如果它被定义为一种亚财产 00:57:14.500 EN: So these sub-property hierarchies also make sense. Usually they are not as large as class hierarchies. Class hierarchies can become quite large and complex. 中文初译: 因此,这些次财产等级也有意义。 通常它们不像阶级等级那么大。 类等级会变得非常庞大和复杂。 00:57:26.500 EN: Property hierarchies are typically not very deep. RDF schema can be used to define a lightweight ontology, but it's not as expressive as OWL, which we will discuss later. 中文初译: 地产等级通常不很深。“RDF”的图案可以用来定义轻量级本体论,但它没有“OWL”那么清晰,我们稍后讨论。 00:57:40.220 EN: And the current standard RDF schema 1.1 has only relatively minor changes over RDF schema 1.0. So it's relatively stable, and you can look at the W3C recommendation page and there you will find the complete definition of RDF schema. 中文初译: 目前的标准“RDF” schema 1. 1 与“RDF” schema 1. 0 相比,只有相对小的改变。 因此它相对稳定, 您可以查看 W3C 推荐页, 这样您就可以找到“ RDF” schema 的完整定义 。 00:58:11.400 EN: So you should be now able to create a small vocabulary, a small schema or ontology. That's why we have such tasks and mini projects here, which you could do. 中文初译: 所以,你现在应该能够创建一个小词汇, 一个小的图案或本体论。这就是为什么我们在这里有这样的任务和小型项目, 你可以做的。 00:58:22.940 EN: So there are some suggestions which you can complete. There are also multiple choice self-assessment tests on SlideWiki, which you can do, practice a bit and prepare for the exam. 中文初译: 所以有些建议可以完成。在幻灯片维基上也有多种选择自我评估测试,你可以做,练习一下,为考试做准备。 00:58:36.860 EN: So for example, one task and mini project is that you create a schema with which you can describe your family situation. 中文初译: 例如,一个任务和小型项目就是 创建一种模式来描述你的家庭状况。 00:58:44.880 EN: You can start with the needed classes and the relations and the first three like spouse, parents, children. Then you can go on with relations in the second degree, sister, brother, grandfather. How would you model gender? 中文初译: 你可以从需要的班级、关系和前三个人开始,比如配偶、父母、子女。然后你可以继续第二等级的关系,姐妹、兄弟、祖父。你会如何模范性别? 00:59:00.080 EN: Male, female, and there are other variations. Meanwhile, as a property of the person's super class or would you create separate classes, man and woman? 中文初译: 男性,女性,还有其他不同。同时,作为个人超级阶级的属性,还是你创造不同的阶级,男人和女人? 00:59:12.650 EN: What are the advantages and disadvantages of this different modeling decisions? What would be better for defining the domain or range of properties like is grandfather off? 中文初译: 这种不同的建模决定的利弊何在? 界定像祖父那样的地产领域或范围更好吗? 00:59:29.460 EN: So this would be one task and I can only encourage you to really develop a small vocabulary using RDF schema because you will understand how the things work in practice by really using this defining classes, subclasses and that could be for example, also a question in the exam later on. 中文初译: 因此,这将是一项任务,我只能鼓励你真正地用“RDF”模式来开发一个小词汇,因为通过真正使用这些定义性的班级、子类,你会理解事物在实践中是如何运作的。 举例来说,这也可以成为以后考试中的一个问题。 00:59:52.140 EN: September 13 will be the exam. We will have it as an online exam. So take the chance. sense to do some of these exercises. 中文初译: 9月13日将是考试,我们将把它作为在线考试。所以,抓住机会,做一些这样的练习。 01:00:07.580 EN: So you could also create a graph representation of different courses presented by different people in the university, for example lectures or seminars as subclasses of courses that are presented by professors or researchers or subclasses of a person. 中文初译: 这样您也可以绘制一个图表,显示大学中不同人士开设的不同课程,例如教授、研究人员或个人分班教授或研究人员讲授的作为次类课程的讲座或研讨会。 01:00:30.230 EN: You can create a turtle or an RDF graph of this example. Here is already an example of that, but you could write this down in turtle for example or create a similar presentation you could create also maybe for your music, maybe exercise what should be modeled as subproperties. 中文初译: 您可以在此创建一个海龟或“ RDF” 图形。 这里已经有一个示例, 但是您也可以在海龟中写下来, 或者创建一个相似的演示文稿, 您也可以为您的音乐创建类似演示文稿, 也许可以练习什么应该做成亚属性模型 。 01:00:54.910 EN: So this is actually not that easy. I gave you this example with Mary 2 being a subproperty of nose actually related to is another property Mary 2 actually could also be a subproperty of being related to, but maybe related to is not a subproperty of nose, not always you know the people you are related to. 中文初译: 所以这其实不是那么容易。我给你们举了这个例子, 玛丽二号是一个鼻子的子财产 实际上与另一件财产有关 玛丽二号实际上也可能是与之有关的次级财产, 但也许它不是鼻子的子财产, 你并不总能了解你的亲属。 01:01:21.030 EN: Some people there are some challenges here also in modeling that. And finally with lists, somehow the lists was missing in the lecture. 中文初译: 有些人在做模型时也遇到一些挑战。最后,随着名单的出现,演讲中遗漏了名单。 01:01:36.550 EN: Maybe we will come back to that next week to lists. I don't know what happened. There were some slides missing about lists, so I will tell you about that next week. 中文初译: 也许我们下周再谈这个问题。我不知道发生了什么。有些幻灯片缺少关于名单的幻灯片,所以我下周再告诉你。 01:01:48.630 EN: Maybe for today I would like to use the chance to show you some further examples of vocabularies. And one of the most common and most popular vocabularies RDF schema, sorry, is schema.org not RDF schema, schema.org. 中文初译: 也许今天我想借此机会向大家展示更多词汇的例子。 而最常见和最受欢迎的词汇“RDF”schema, 抱歉, Schema.org不是“RDF”schema, schema.org。 01:02:08.430 EN: You can go to the web page schema.org, which you can see here. And it looks very modest, very simple web page, but it's actually done by Google, Microsoft, Pinterest, Yandex, and many other web and e-commerce companies. 中文初译: 你可以去Schema.org网页,这里可以看到。它看起来很谦虚,非常简单, 但实际上是由谷歌、微软、Pinterest、Yandex和其他许多网络和电子商务公司完成的。 01:02:29.410 EN: And it defines schemas for publishing data on the web. So basically RDF schemas. It's also run basically organized on GitHub. 中文初译: 它定义了在网络上发布数据的模式。 基本来说, “ RDF” 模式。 它也基本上以 GitHub 为主。 01:02:40.750 EN: But when you click on schemas, you basically get an overview of the schemas. And here you can see what kind of schemas are defined. 中文初译: 但是当你点击Schemas时, 你基本上可以得到一个对schemas的概览。在这里你可以看到什么是 shomas 的定义。 01:02:51.970 EN: So these are basically classes which are defined. They call it schemas, but it's basically each schema is what we would call in RDF schema a class. 中文初译: 这些基本上都是定义的类别。他们称之为“计划”, 但基本上每个方案都是我们所谓的“RDF”“计划”类。 01:03:02.830 EN: So and they have defined here, for example, creative work, book, movie, music, recording, recipe, TV series. Let's maybe click at book. 中文初译: 因此他们在这里定义了创意作品, 书籍,电影,音乐,录音,食谱,电视系列。让我们点击书本吧。 01:03:13.310 EN: And when we click at book, we see here on top the hierarchy. So book is a subclass of creative work and is a subclass of thing on top. 中文初译: 当我们点击书本时,我们可以看到这里的等级最高层。所以书是创造性作品的一个子类, 上面是事物的一个子类。 01:03:24.630 EN: Let me increase the size here a bit that you can see this better. Maybe like this. So here you can actually see the subclass relationships there on top. 中文初译: 让我在这里增加一些大小, 以便您能够看得更清楚。 也许像这样。 所以这里你可以看到上面的子类关系。 01:03:36.170 EN: And what we see below is the properties which have book as a domain. So these properties and we see the range of the properties. 中文初译: 下面我们看到的是将书书作为域的属性。 因此这些属性和我们可以看到属性的范围 。 01:03:46.050 EN: They call it your expected type. So don't be confused. They organized it a bit simpler so that it's easy to understand. 中文初译: 所以别搞糊涂了,他们组织得简单一点 这样很容易理解 01:03:52.970 EN: So the properties which have book as a domain are, for example, abridged, book addition, book format, illustrator, ISBN. 中文初译: 因此,以书籍为域的属性是,例如,简略、书添加、书格式、插图、ISBN。 01:04:04.170 EN: ISBN is a very important property, right? So this has book as a domain. And it has as a range. ISBN has as a range just text here. 中文初译: ISBN 是一个非常重要的属性, 对不对? 因此它有书作为域。 它有范围。 ISBN 这里只有范围文字 。 01:04:14.470 EN: So schema.org is also relatively simple. It's maybe not the best schema or from the knowledge representation point of view. 中文初译: Schema. org 也相对简单。它也许不是最好的方法, 或者从知识的表达角度来说。 01:04:22.610 EN: For example, ISBN you would model differently as a number and you would create maybe your own data type for it. 中文初译: 例如, ISBN 您会以不同的数字模式建模, 然后为它创建您自己的数据类型。 01:04:31.670 EN: Here they kept it simple and it's just the text. But number of pages, for example, is modeled as an integer value. 中文初译: 他们在这里保持简单,只是文字而已。但是,例如,页数是按整数值建模的。 01:04:41.010 EN: So the number of pages in the book. So these are actually the properties which have book as domain. And then below we see properties from creative work. 中文初译: 书中的页面数。 因此这些实际上是将书籍作为域名的属性。 然后在下面我们可以看到创造性作品的属性 。 01:04:53.230 EN: So these are the properties which have creative work as a domain. Since book is a subclass of creative work, all the properties from creative work also apply to book. 中文初译: 这些属性是具有创造性作品的域名。由于书籍是创作作品的子类,因此所有创意作品的特性也都适用于图书。 01:05:07.450 EN: We can apply all these properties which are defined for creative work also for books. And then we have, for example, about. 中文初译: 我们可以应用所有这些特性,这些特性被定义为创造性作品, 也用于书籍。然后我们有,比如说,关于。 01:05:14.950 EN: So the subject matter of the content of the book or the creative work. There can be an abstract and access mode here. 中文初译: 因此该书内容或创作作品的主题事项。 这里可以有一个抽象的存取模式 。 01:05:25.230 EN: So you see it's a long list of properties. An audio, the audience defined, citation, comments, count of comments, content rating, how it's rated by people, the contributor who contributed to this creative work or here then the book copyright and so on and so forth. 中文初译: 所以,你看这是一个很长的属性列表。 音频、 读者定义、 引用、 评论、 评论、 评论、 内容评分、 人们如何评分、 贡献者对创作工作的贡献, 或者在这里, 然后是书版权等等。 01:05:56.690 EN: So they are alphabetically ordered here and you can see there are a lot of properties meanwhile defined. And in the rightmost column we have a description. 中文初译: 因此它们按字母顺序排列在这里, 您可以看到同时定义了许多属性。 在最右侧的列中, 我们有一个描述 。 01:06:07.960 EN: So basically this is a description of the property. So indicates whether the book is an abridged edition. So this would be an RDF schema, the RDF description property. 中文初译: 基本上,这是对财产的描述。 所以表示这本书是否是简略版。 因此这将是一个“ RDF” 模型, “ RDF” 描述属性。 01:06:27.340 EN: So you can see here basically this was book. Let's maybe look at some other examples. So there are other creative works like movies as well. 中文初译: 所以,你可以在这里看到,基本上这是一本书。我们也许可以看看其他的例子。所以还有其他创造性作品,比如电影。 01:06:38.600 EN: Maybe let's have a look at movie. Again here we have the properties which are defined specifically for movie. And on top we again see that movie is a subclass of creative work. 中文初译: 也许让我们看看电影吧。再次,我们这里有专门为电影定义的属性。 最重要的是,我们再次看到这部电影是创造性作品的一个子类。 01:06:51.420 EN: And a movie of course has actors and that's why it has a property actor and the expected type. So the range of the property actor is a person. 中文初译: 电影当然有演员,这就是为什么它有一个房地产演员和预期的类型。所以,房产演员的范围是一个人。 01:07:03.080 EN: And it has a country of origin and the range is a country. And you see you can actually also click here when you click on person then we actually get to the class description of person. 中文初译: 它有一个起源国, 范围是一个国家。 你看您也可以点击这里, 当您点击一个人时, 然后我们实际上进入了个人分类描述 。 01:07:14.740 EN: So person is a direct subclass of thing here in that case. And then we have properties defined for person like address of a person affiliation. 中文初译: 人就是直接的子类。然后,我们为一个人定下了财产,比如个人属性的地址。 01:07:28.800 EN: Person can be alumni of an educational organization. It can have gotten an award as a birthday, birthplace and so on and so forth. 中文初译: 个人可以是一个教育机构的校友,也可以作为一个生日、出生地等等获得奖项。 01:07:43.000 EN: So a long list of properties. And you also see there are some properties which are already defined for the super class of everything. In that case here thing. 中文初译: 如此长的属性列表。 您也可以看到有些属性已经被定义为所有事物的超级类。 在这种情况下, 这里有些东西 。 01:07:55.020 EN: So and there's some of these generic properties from thing like an identifier, a description, an alternate name, an image. 中文初译: 所以有些通用属性来自 标识符、描述、替代名称、图像等。 01:08:05.280 EN: That's something which can be applied to everything in schema.org. That's why these properties are already defined or have as a domain thing. 中文初译: 所以这些属性已经被定义为或具有域名。 01:08:20.630 EN: So schema.org is a very important vocabulary. So if you search on the web for people for example. Let's search for Angela Merkel. 中文初译: 所以 schema. org 是一个非常重要的词汇。 因此如果你在网上搜索人们, 比如说。 让我们搜索安格拉·默克尔。 01:08:36.400 EN: Then you see on the right hand side here in your web browser such a fact box or in the search. So you see an image, you see when she was born, how tall she is, what party she is a member of, who is her husband, her alumni here, the educational institution here. 中文初译: 然后,你可以看到右手边在你的网络浏览器里 像这样的事实框或搜索中。你看一个图像, 你看她出生时,多高, 她是个什么样的政党的成员, 她的丈夫是谁,她的校友在这里,这里的教育机构。 01:08:59.660 EN: And that actually comes from a knowledge graph which uses schema.org. The schema.org vocabulary. Of course for Angela Merkel, Google is creating its knowledge graph. 中文初译: 这其实来自一个使用 schema.org 的知识图。schema.org 词汇表。当然,对于Angela Merkel来说,谷歌正在创建其知识图。 01:09:11.920 EN: But you can also search for more specific things like small restaurants. I don't know. Let's search for a small restaurant which probably Google has not added himself. 中文初译: 但您也可以寻找更具体的东西, 比如小餐厅。我不知道。让我们去寻找一个小餐馆, 谷歌可能没有加到自己身上。 01:09:29.100 EN: So here is a small restaurant in Leipzig. You have an address here. You see the opening hours. You see I have been there also four years ago. 中文初译: 这里是莱比锡的一家小餐厅,这里有地址。你看营业时间,四年前我也去过那里。 01:09:39.500 EN: You have a description here of the restaurant. You have the reviews 365 ratings. You have then an aggregated review count. You have the phone number. 中文初译: 您在这里有餐厅的描述。 您得到了365个评分。 然后是综合评分计数。 您有电话号码 。 01:09:54.440 EN: And these are all things which come from schema.org. And Google didn't add this manually, but what they did is they basically extracted it from the website from this restaurant. 中文初译: 这些都是来自schema.org。Google没有手工添加这个内容, 但是他们所做的就是从这家餐厅的网站中提取了它。 01:10:07.340 EN: And let's maybe have a look how this works. So we go, sorry to schemas. Let's look at restaurants. So here we have restaurant as a class. 中文初译: 让我们来看看这是怎么回事。所以我们去,抱歉到Schemas。让我们来看看餐馆。这里我们有一个作为班级的餐厅。 01:10:19.760 EN: And here you can see exactly these properties. First on top you see to what classes it belongs. So restaurant belongs to food establishment, to local business, to organization, and then to thing. 中文初译: 这里您可以看到这些属性。 首先,你看看它属于哪一类。 所以餐馆属于食品机构, 当地企业,组织,然后是事物。 01:10:33.180 EN: It's a subclass of food establishment. Food establishment is a subclass of local business. Local business is subclass of organization. 中文初译: 食品机构是当地企业的子类,地方企业是组织组织的子类。 01:10:41.140 EN: But it also belongs to another path in the class hierarchy. It belongs also to food establishment, to local business, but local businesses are also places. 中文初译: 但是,它也属于阶级等级制度的另一条道路。 它也属于食品机构、当地企业,但地方企业也是场所。 01:10:52.040 EN: So here you see this multiple inheritance. A class, in that case local business, is a subclass of both organization and place. 中文初译: 所以这里你可以看到这种多重继承。一个班级,如果是当地企业, 是组织和地点的子类。 01:11:01.240 EN: And that's why it inherits the properties. So the properties which are defined for organization can be applied for local businesses, but also the properties which are defined for place can be applied to local businesses and then to restaurants. 中文初译: 这就是为什么它继承了财产。因此,为组织而定义的财产可以适用于当地企业,但为地方而定义的财产也可以适用于当地企业,然后适用于餐馆。 01:11:17.620 EN: And then we see here the properties from food establishment. Then the properties from local business. Basically you go up in the hierarchy and you look at all properties which have the domain of these super classes. 中文初译: 然后我们从这里看到食物的产物。然后是当地企业的产物。基本上你进入等级体系, 并查看所有具有这些超级等级范围的属性。 01:11:31.920 EN: And then they are listed here nicely on one page. So we have the properties also from organization. An organization has an address. It has an alumni. 中文初译: 然后在一页上列出它们。所以我们也有组织的属性。一个组织有一个地址,它有一个校友。 01:11:45.660 EN: And here you see already also the bit the problems of schema.org and alumni for a restaurant doesn't make so much sense, but can be actually defined formally. So schema.org is, but it was also criticized a bit that there are some problems in there. 中文初译: 这里你已经看到Schema.org和餐馆校友的问题 并不合理,但实际上可以正式定义。Schema.org是,但也有人批评说那里有些问题。 01:12:04.180 EN: For example, there was one phone number for places. You have defined phone number. And then one special type of place is a mountain. 中文初译: 例如, 地方只有一个电话号码。 您定义了电话号码。 然后一种特殊的地方是山。 01:12:15.280 EN: And then a mountain has a phone number. So there are things, some things in schema.org which maybe are a bit questionable from a knowledge representation point of view. 中文初译: 然后一座山就有一个电话号码。所以有些东西, 在 schema.org 中有些东西, 从知识的表达角度来说,这些也许有点问题。 01:12:25.740 EN: So it's a simplification, but it still works relatively well. And it's used on around 30, 40% of the web pages. 中文初译: 简化了一点,但效果还是比较好。大约30到40%的网页都使用了它。 01:12:36.020 EN: And that I wanted to show you as well. So if you scroll down here on schema.org, you can then see examples down there. 中文初译: 我想也给你看。所以如果你在 schema.org上滚下来, 你可以在那里看到一些例子。 01:12:44.720 EN: So you can see an example here of a restaurant, how it looks as a text. Yeah, we have the name of the restaurant. 中文初译: 因此,你可以在这里看到一个餐馆的例子, 它如何看成文字。是的,我们有餐厅的名称。 01:12:52.520 EN: It has four stars based on 250 reviews. You have the address. You have a link to the web page here. You have the opening hours. 中文初译: 它有四颗星基于250个评论。 您有地址。 您可以链接到此网页。 您有开放时间 。 01:13:02.840 EN: You have the categories. So it serves Mediterranean food here, the price range, and whether it accepts reservations. 中文初译: 你有这些类别。所以它在这里为地中海食物服务, 价格范围,以及它是否接受保留。 01:13:11.140 EN: If it would be just as text or as HTML, the search engine would not understand. For example, the opening hours here, it's sometimes difficult to parse and to understand somebody. 中文初译: 如果它和文本或 HTML 一样, 搜索引擎将无法理解。 例如, 这里的打开时间, 有时很难解析和理解某人 。 01:13:24.120 EN: For example, in Europe, you don't use AM and PM. You use different opening hours on the web page. So Google could not understand when a restaurant would open if because in every region, it would be done a bit differently. 中文初译: 例如,在欧洲,您不使用AM和PM。您在网页上使用不同的开放时间。因此谷歌无法理解餐厅何时开业,因为在每个区域,它会做一些不同的事情。 01:13:41.800 EN: That's why the web page is annotated with RDF. And that's RDF-A. We will explain. Maybe did we explain to you? I think Marcos explained it maybe to you how RDF-A works. 中文初译: 这就是为什么网页上加了“RDF”的注解。这是“RDF-A ” 我们来解释一下。也许我们向你解释过吗?我想Marcos向你们解释了“RDF-A”是如何运作的。 01:13:58.520 EN: And there you actually embed this information in HTML. Yeah. And here you can see how this looks. So we have, in addition to the HTML text, we have the RDF-A annotation. 中文初译: 你实际上将这些信息嵌入到 HTML 中。 是的。 这里你可以看到这个样子。 因此,除了 HTML 文本之外, 我们还有“ RDF ” 注释。 01:14:13.380 EN: And we have, for example, here indicated the property that it's the name of the restaurant. And then Google understands, ah, this is the name. 中文初译: 例如,我们在这里指出了餐厅的名称。然后谷歌理解了, 啊,这就是它的名字。 01:14:21.300 EN: We have the aggregated rating here. And then we have the value aggregated rating value of four. And then we have the review count of 250. 中文初译: 我们这里有总评级。然后我们有四倍的总评级值。 然后我们有250个审查数。 01:14:31.840 EN: So then Google and other search engines can actually extract this information from the web pages and can present it in the way as we see it on the web page. And that's exactly how that works. 中文初译: 谷歌和其他搜索引擎可以从网页上提取这些信息, 并以我们在网页上看到的方式展示这些信息。 这正是这个方法的原理。 01:14:44.940 EN: And that's done by millions of web pages because, of course, all webmasters try to do search engine optimization to organize the content in such a way that it can be easily discovered by search engines. 中文初译: 这是数百万网页完成的, 因为,当然,所有网络技术主管 都试图做搜索引擎优化 来组织内容, 以便很容易被搜索引擎发现。 01:15:01.140 EN: And that's why around 30 percent of the web pages contain the schema.org markup. Good. And that's an example of the maybe one of the largest and most popular vocabularies. 中文初译: 这就是为什么大约30%的网页 含有 schema.org 标记。很好,这是可能最大的和最受欢迎的词汇之一的例子。 01:15:20.080 EN: Let's maybe let me show you another one that's friend of a friend. For the full vocabulary. That's one of the oldest vocabularies, one which was created initially. 中文初译: 让我再给你们看一个朋友的朋友, 完整的词汇表。这是最古老的词汇表之一, 最初创建的。 01:15:35.220 EN: And here. You can see the definition of this vocabulary. So it looks a bit like a standard document. You have a license and the authors also here Creative Commons license. 中文初译: 这里,您可以看到这个词汇的定义。 所以它看起来有点像标准文件。 您有许可证, 作者也在这里使用创意共同公司许可证。 01:15:51.460 EN: You have a summary. But then basically it consists of classes and property definitions. And that's what we see here a bit down there. 中文初译: 你有一个摘要,但基本上它包含类别和财产定义。这就是我们下面看到的。 01:16:03.800 EN: We have classes and properties. For example, we also have person. We can click on person and then we have a short definition here, description. 中文初译: 我们有分类和属性。例如,我们也有个人。我们可以点击一个人,然后我们有一个简短的定义,描述。 01:16:15.140 EN: Here's actually a longer description. And then we have the properties which have person as a domain. So surname, last name, family name, publications, work info, home page and so on. 中文初译: 这里实际上有一个较长的描述。 然后我们有了将个人作为域名的属性。 因此, 姓氏、 姓、 姓氏、 出版物、 工作信息、 主页等等 。 01:16:30.080 EN: These are properties which can be used with person. And it's subclass of agent and spatial thing. And it's disjoint with a project and organization. 中文初译: 这些属性可以与人一起使用。 并且是代理人和空间事物的子类。 它与一个项目和组织脱节。 01:16:48.400 EN: So we can also click on the properties, for example, nose here. And then we have a definition for the nose property. 中文初译: 所以我们也可以点击这里的鼻子等属性。 然后我们有了鼻子属性的定义 。 01:16:55.920 EN: A person known by this person. So it indicates some level of reciprocated interaction between the parties, between two people. That's a bit philosophical definition of knowing. 中文初译: 这个人认识的人。它表明双方之间, 两个人之间 某种程度的对等互动。 这是一个有点哲学意义上的认知定义。 01:17:12.640 EN: And nose has domain person and as range person. And then there is a longer explanation here. It's actually quite extensive. 中文初译: 鼻子有域名和射程。然后这里有一个更长的解释。它实际上非常广泛。 01:17:24.060 EN: And you can here browse this vocabulary and get information about this friend of a friend vocabulary. And there are many of these vocabularies. 中文初译: 您可以在这里浏览这个词汇, 并获得关于朋友词汇中这位朋友的信息。 而且有很多这样的词汇表。 01:17:34.940 EN: So friend of a friend is something which represents information about social networks. Actually, some social networks also use this vocabulary to publish information or to make information available according to the FOA vocabulary. 中文初译: 所以朋友的朋友是代表社交网络信息的东西。事实上,一些社交网络也使用这个词汇来发布信息或根据FOA词汇提供信息。 01:17:49.880 EN: And meanwhile, there are many of these vocabularies which use RDF schema, basically, and also OWL. And I want to show you one registry for these vocabularies, LOV, linked open vocabularies. 中文初译: 同时,这些词汇中有许多使用“RDF”的模型,基本上,还有“OWL”。我想给你们展示一个这样的词汇登记册,LOV,连接起来的开放词汇。 01:18:10.100 EN: And there you can search for vocabularies. So you see here the most common used ones. You basically have a cloud of vocabularies. 中文初译: 在那里,你可以搜索词汇表。所以你可以看到这里最常用的词汇表。基本上你有一个词汇表云。 01:18:21.560 EN: You can then browse them, but you can search for someone. For example, if you look for vocabularies about product. 中文初译: 您可以浏览它们, 但是您可以搜索某个人。 例如, 如果您要寻找关于产品的词汇 。 01:18:31.510 EN: So you can look at product and then find vocabularies which define product. So here you see BBC has defined product. 中文初译: 所以您可以查看产品,然后找到定义产品的词汇表。在这里,您可以看到英国广播公司已经定义了产品。 01:18:48.070 EN: BBC, that's the British Broadcasting Corporation, so a media agency. Of course, their product definition is, I guess, movies, documentaries, TV programs. 中文初译: 英国广播公司,英国广播公司,所以是一个媒体机构。当然,他们的产品定义是,我猜,电影,纪录片,电视节目。 01:19:03.270 EN: So that's why BBC might not be something which you would use in an industrial context. And you can also look then for other vocabularies which are maybe more suitable for industrial products. 中文初译: 这就是为什么BBC可能不是你们在工业环境中会使用的东西。你也可以去寻找其他的词汇,这些词汇也许更适合工业产品。 01:19:21.670 EN: We could, for example, look at that one here. That's a very short definition. So you can basically click. In schema.org, there is also product. So that's linked here. 中文初译: 例如,我们可以在这里看看这个。这是一个非常简短的定义。 所以基本上你可以点击。 在 schema. org 中, 也有产品。 因此它在这里是相连的 。 01:19:40.560 EN: So you can basically browse through vocabularies and search for properties. Here is one for software product. And if you click on that, you basically get a definition of this product. 中文初译: 因此,您可以浏览词汇和搜索属性。这里是用于软件产品的软件。如果你点击这个,你基本上可以得到该产品的定义。 01:19:55.060 EN: Here that's from the German National Library. So they have also defined a large ontology as well. Good. Maybe that's all for today. Thanks for your attention. 中文初译: 这儿是德国国家图书馆的。所以他们也定义了一个巨大的本体论。很好,也许今天就到此为止了。谢谢大家的注意。 01:20:11.930 EN: And we have time for some questions. 中文初译: 我们还有时间回答一些问题。