review1

1. XML Namespaces

1.1 Naming conflicts exist in different XML documents

XML document_1
<table>
<tr>
	<td>Apples</td>
    <td>Bananas</td></ tr>
</table>

XML document_2
<table>
	<name>African Coffee Table</name>
    <width>8G< /width>
<length>120</length>
< /table>

1.2 Solution : Adding prefixes as namespaces

XML document_1
<h:table>
	<h:tr>
		<h:td>Apples</h:td>
        <h:td>Bananas</h:td>
    </h:tr>
</h:table>

XML document_2
< f:table>
	<f:name >African Coffee lable</f:name>
    <f:width>80</f:width>
	<f:length>120</f:length>
</f:table>

1.3 XML Namespace Syntax:

URI 英文字符 IRI任何编码

2. RDF

the data model of Semantic Technologies and of the Semantic Web;
- structures metadata about Web sites, pages, etc.:
  1. Page author, creator, publisher, editor,…
  2. Data about them: email,phone, job,…
can be used for machine-readable data exchange;
is introduced in W3C Recommendation of 1999 (Version 1.O);- uses XML as main syntax for serialization.

ARDF knowledge base: a directed labeled graph, i.e., Knowledge Graph

2.1 RDF Triple (Statement)

Subjects: Resource or blank node
Predicates: Resource
Object: Resource, literal or blank node

Literals are:

data values;
encoded as strings;
interpreted by datatypes;
treated the same as strings without datatypes, called plain literal;
- A plain literal may have a language tag;
- Datatypes are not defined by RDF, but usually from XML Schema.

Typed Literals:

“Beantown3’^^xsd:string“

The Bay State”^^xsd:string
Plain literal and literals with language tags:

“France”‘France”@en

“France”‘@fr
“法国”@zh“

“Frankreich”@de

Equalities for Literals:
“0013^^xsd:integer =“13”^^xsd:integer
“123.0^^xsd:decimal= “00123”1^^xsd:integer (based on data type hierarchy)

Exercise

Does the datatype“德国”equals to“德国”@ zh ?

不一样

Blank node: unnamed resource or complex node

The set of blank nodes, the set of all lRls (named resources)and the set of all literals are pairwise disjoint;
Representation of blank nodes is syntax-dependent:underline+colon+ID (Turtle syntax):_:xyz,_:bn;
The scope of the lD of a blank node is only the document towhich it belongs.

2.2 RDF: N-ary Relations

Example:

1
2
3

prefix ex: <http://lexample.org/> .
ex: Chutneyex:hasIngredient "1lb green mango",
						"1tsp. Cayenne pepper" .

@prefix ex :<http://lexample.orgl> .
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
ex:Chxtney ex:hasIngredient ex:ingredient1 . 
ex:ingredient1 rdf:value ex:greenMango;
			   ex:amount  "11b" .

<rdf:Description rdf:about="http://example.org/Chutney">
	<ex:hasIngredignt rdf:nodeID="id1"/>
</rdf:Description>
<rdf:Description rdf:nodeID="id1">
	<ex:ingredient rdf:resource="http://example.org/greenMango"/>
	<ex:amount>1lb</ex:amount>
</rdf:Description>

<rdf:Description rdf:about="http://example.org/Chutney">
	<ex:hasIngredient rdf:parseType="Resource">
		<ex:ingredient rdf:resource="http://example.org/greenMango"/>
        <ex:amount>1lb</ex:amount>
	</ex:hasIngredient>
</rdf:Description>

sw:John sw:is_a sw:professors ;
		sw:has_name "John Doe”;
		sw:has_id “987654321".
sw:John sw:is_a sw:professors .
sw:John sw:has_name “John Doe”.
sw:John sw:has_id 987654321.

3. RDFS

provides a data-modeling vocabulary for RDF data;
is an extension of the basic RDF vocabulary;
tries to provide consistent explanations for RDF data;- allows for specifying schema knowledge;
Mothers are female
Only persons write books
is a part of the W3C Recommendation.

Why do we need RDFS when there already exists XML Schema?
- XML Schema only defines syntax without semantics;
- XML Schema cannot reference “things”outside the document.

3.1 RDFS: Class and Instance

Given a triple:

1	ex:SemanticWeb rdf:type ex:Textbook .

which characterizes“Foundations of Semantic Web Technologies”as aninstance of the class “Textbook”.

Aresource can be the member of more than one class

1	ex:SemanticWeb rdf:type ex:Book .

lnstance and class names cannot be distinguished syntactically with lRls..
RDFS helps explicitly state that a resource denotes a class:
1
ex:book rdf:type rdfs:Class .
rdfs:Class is the“class of all classes” .

3.2 Class Hierarchy (Taxonomy)

3.3 Property and Property Hierarchy

Allow to state that a certain property can only be between things of a certain class:
- e.g. when a is married to b, then both a and b are Persons

1 2	ex:isMarriedTo rdfs:domain ex:Person. ex:isMarriedTo rdfs:range ex:Person

Datatypes can also be used for adding property restrictions:

1	ex:hasAge rdfs:range xsd:nonNegativeInteger

3.4 RDFS: Reification

How to graphically represent a sentence by means of the blank node?

3.5 Example: Reasoning with RDFS

4. Constituents of a DL Knowledge Base

4.1 Description Logic: ALC

ALC: the simplest DL

complex concepts are defined as follows:

⊥ and T are concepts；
For concepts C and D, ¬C, C⊓D, and C⊔D are concepts;
For a role r and a concept C, ∃r.C and ∀r.C are concepts

Example: Student ⊓ ∀attendsCourse.MasterCourse
- It describes the concept comprising : all students that attend only master courses.

4.2 Example: Are the following valid DL concepts

𝐴∀𝑅.𝐵
(𝐴⨅𝐵)⨆𝐶
(∃𝑅.𝐴)⨅𝐵
¬𝐴⨆𝐶¬

×√√×

4.2.1 TBox:

For concepts C, D, a general concept inclusion (GCI) axiom has the form: C ⊑ D
1. C≡D is an abbreviation for C ⊑ D and D ⊑ C.
2. a TBox (terminological Box) consists of a set of GCIs.

4.2.2 ABox:

an ALC ABox assertion can be of one of the following forms:
1. C(a), called concept assertion
2. r(a, b), called role assertion - an ABox consists of a set of ABox assertions

(1) Any student must attend at least one lecture.
Student ⊑ ∃Attend.Lecture

Students are those who attend at least one lecture?

不对的，因为这个are表示是完全相等的关系，但是后面那个concept的范围更大

(2) John is a brother of Mary.
Brother(John,Mary)
(3) Parents are exactly those who are mothers or fathers.
Parent ≡ Mother ⊔ Father

A set of axioms (knowledge base) is satisfiable (or consistent) if it has a model.
It is unsatisfiable (inconsistent) if it does not have a model.
Inconsistency is often caused by modeling errors.
- Unicorn ⊑ Fictitious
- Unicorn ⊑ Animal
- Animal ⊑¬Fictitious

5. Deductive Reasoning: Backward Reasoning

Backward reasoning (Backward chaining) is an inference method described colloquially as working backward from the goal.
It is often used in entailment checking or query answering in KG.
It uses the rules to rewrite the query in several ways and the initial query is entailed if a rewritten query maps to the initial facts.

Example1:
Rule: If you are human, then you are mortal.

Human(x)→Mortal(x)
Question: Is Socrates a mortal?

Solution: Check whether Mortal(Scorates) or Human(Scorates) is true.

5.1 Other Tasks in Logical Reasoning: Inconsistency Checking

Incoherent ontology: ontology with at least one unsatisfiable concept. Example: PhDStudent ⊑ Student, PhDStudent ⊑ Employee, Student ⊑Employee
Inconsistent ontology: ontology without a model. Example: PhDStudent ⊑ Student, PhDStudent ⊑ Employee, Student ⊑Employee, PhDStudent(John)

6. TransE: Take Relation as Translation

For a fact (head, relation, tail), take the relation as a translation operator from the head to the tail .

TransE
- For each triple , h is translated to t by r.

Question

7. RDB2RDF

Direct Mapping

7.1 RDB2RDF: R2RML

R2RML is a language for specifying mappings from relational to RDF data.
A mapping takes as input a logical table, i.e.,
a database table
a database view (a virtual table collecting data from relational tables), or
an SQL query (called an “R2RML view” because it is like an SQL view but does not modify the database)

A triples map has three parts: - the input logical table
a subject map
several predicate-object maps (combining predicate and object maps).

8. Exercise

Please represent the following sentences graphically by means of blank nodes.
Jonathon says cherries taste sweet.
1月14日，国家药监局宣布酚酞片停止生产.

What can be inferred from the following triples using RDFS semantics?

ex:Undergraduate_student rdfs:subClassOf ex:Student .
ex:Postgraduate_student rdfs:subClassOf ex:Student .
ex:Professor rdfs:subClassOf ex: Academic_staff .
ex:Academic_staff rdfs:subClassOf ex:University_staff .
ex:Teach rdfs:domain ex:Academic_staff .
ex:Teach rdfs:range ex:Student .
ex:Supervise rdfs:domain ex:Professor .
ex:Supervise rdfs:range ex:Postgraduate_student .
ex:John ex:Supervise ex:Mary .

ex:Professor rdfs:subClassOf ex: University_staff .
ex:John rdf:type ex:Professor .
ex:Mary rdf:type ex:Postgraudate_student .
ex:John rdf:type ex:Academic_staff .
ex:John rdf:type ex:University_staff .
ex:Mary rdf:type ex:Student .

Express the following sentences in Description Logics.

Li is a student of SEU.
Any person who is enrolled in a university is a college student.

Are the following valid OWL axioms? If not, why? (class: A,B,C,D, role: R)

A⨅(B⨆ ∃R)
(∀R.A⨆B)⨅A
(A⨅B)¬C⨆D

5) Please write the corresponding RDF triples (in Turtle) of the following sentence: “Grandpas are the men who are grandparents”. Specific classes are defined in the namespace
http://semanticweb.org/. Prefixes are given as follows:
@prefix sw: http://semanticweb.org/ .
@prefix rdf: http://www.w3.org/1999/02/22-rdf-syntax-ns# .
@prefix owl: http://www.w3.org/2002/07/owl# .

1
2
3

_:x  rdf:type  owl:class  ;
     owl:intersectionOf  (sw:Man  sw:Grandparent)  ;
     owl:equivalentClass  sw:Grandpa .

Give a model of the following ontology

Professor⊑FacultyMember,
Professor(Guilin Qi),
hasAffiliation(Guilin Qi, SEU),
StudentOf(Zhang, Guilin Qi)

A model:

∆={a,b,c}
I(Guilin Qi)=a, I(SEU)=b, I(Zhang)=c
I( Professor)={a}, I(FacultyMember)={a},
I(hasAffiliation)={(a,b)},
I(StudentOf)={(c,a)},

Given the following KG, use query rewriting on the query Person(x) (this query is used for checking whether x is a person) .

Given a relational database and RDF triples, please write corresponding R2RML triples maps which can map the given RDB to RDF.

<http://data.example.com/student/001> rdf:type ex:Student ;
                                                                ex:id "001" ;
                                                                ex:name "Wang" ;
                                                                ex:major <http://data.example.com/major/211> .
<http://data.example.com/student/002> rdf:type ex:Student ;
                                                                ex:id "002" ;
                                                                ex:name "Wu" .
<http://data.example.com/major/211> rdf:type ex:Major ;
                                                             ex:name "AI" .