=Paper=
{{Paper
|id=Vol-2293/paos2018-passcr2018_paper1
|storemode=property
|title=
Building up Ontologies with Property Axioms from Wikipedia
|pdfUrl=https://ceur-ws.org/Vol-2293/paos2018-passcr2018_paper1.pdf
|volume=Vol-2293
|authors=Tokio Kawakami,Takeshi Morita,Takahira Yamaguchi
|dblpUrl=https://dblp.org/rec/conf/jist/Kawakami0Y18
}}
==
Building up Ontologies with Property Axioms from Wikipedia
==
https://ceur-ws.org/Vol-2293/paos2018-passcr2018_paper1.pdf
Building up Ontologies with Property Axioms
from Wikipedia
Tokio Kawakami1 , Takeshi Morita1 , and Takahira Yamaguchi1
1
Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama, Kanagawa 223-8522, Japan
kawakami0412@keio.jp,{t morita,yamaguti}@ae.keio.ac.jp
Abstract. Wikipedia has been recently drawing attention as a semistruc-
tured information resource for the automatic construction of ontology.
We proposed a method of constructing a general-purpose ontology by au-
tomatically extracting the is-a relations (rdfs:subClassOf), class-instance
relations (rdf:type), property relations and types. Methods to automati-
cally construct ontologies with numerous property relations and types
from Wikipedia are discussed here. The ontologies include is-a rela-
tions, class-instance relations and property relations and types. The
property relations are triples, property domain (rdfs:domain), property
range (rdfs:range), and property hypernymy-hyponymy relations (rdfs:su
bPropertyOf). The property types are object (owl:ObjectProperty), data
(owl:DatatypeProperty), symmetric (owl:SymmetricProperty), transitive
(owl:TransitiveProperty), functional (owl:FunctionalProperty), and in-
verse functional (owl:InverseFunctionalProperty).
Keywords: Ontology Learning · Property Axioms · Wikipedia
1 Introduction
It is effective to construct large-scale ontologies for information searching and
data integration. Among popular ontologies are WordNet [1] and Cyc [2]. How-
ever, it is expensive to construct these ontologies manually. Moreover, the manual
ontology engineering process results in numerous bugs, and its maintenance and
update are challenging. Therefore, more attention comes to build the automatic
or semi-automatic creation of ontologies on research, ontology learning.
Wikipedia, the web-based open encyclopedia, is increasing in popularity as
a new information resource. Because Wikipedia has rich vocabulary, reasonable
updatability, and semistructuredness, there is less differences between Wikipedia
and ontologies when compared with free text. Thus, ontology learning from
Wikipedia is becoming popular.
We proposed a large-scale and general-purpose ontology learning method
for extracting the is-a relations (rdfs:subClassOf), class-instance (rdf:type), and
property relations and types by using Wikipedia as the resources [3]. However,
there are certain challenges. For example, we did not define the property domain,
propety range and types. Therefore, we extract the property names (that are
including following types: object, data, symmetric, transitive, functional, and
2 T. Kawakami et al.
inverse functional) and relations (that are triples, property domain and property
range) by adding certain techniques to current technique and construct a large-
scale and general-purpose ontology including numerous properties.
This paper is structured as follows: We introduce related works on deriving
ontology from Wikipedia in Section 2. In Section 3, we explain the definition of
property as described in our previous research and details on extraction tech-
niques applied to Wikipedia. In Section 4, we present the result of the experiment
wherein we applied the extraction techniques to Wikipedia. Finally, we present
the conclusion of this paper and our future work.
2 Related Work
Auer et al.’s DBpedia [4] constructed a large information database to extract
RDF from semi-structured information resources of Wikipedia. They used in-
formation resources such as Infobox, external link, categories the article belongs
to. However, properties and classes are constructed manually, and there are 170
classes and 720 properties. We use not only Infobox and Wikipedia categories
but also text information such as list structures, list articles and definition texts
and extract relations automatically. It is also different in that it is working on
automatic extraction of property axioms.
Fabian et al. [5] proposed YAGO, which enhanced WordNet by using the
Conceptual Category. Conceptual Category is a category of Wikipedia English,
whose head part has the form of a plural noun suchas American singers of
German origin . They defined a Conceptual Category as a class and defined the
articles that belong to the Conceptual Category as instances of the class. This
method permits all articles using categories to be set as an instance. However,
as the information in the main text is not used, in the cases where the article
is absent or where information not reflected in the categories is present in the
main text, it is unfeasible to extract such information as the instance.
YAGO2 [6] aimed for further expansion of ontology with the knowledge
base expansion of YAGO both by the previous linkage between WordNet and
Wikipedia category and by the extraction of spatiotemporal information from
Wikipedia and GeoNames3. Moreover, in the expansion version, YAGO3 [7],
Wikipedia (in other languages as well as English) is used to permit multilin-
gual expansion. YAGO2 focused on nonhierarchical relations and constructed
an advanced ontology, incorporating a spatiotemporal relation that is not based
only on the hierarchical relation. However, it has not utilized Wikipedia’s unique
structural aspects such as information in the main text, define statements, and
Wikipedia lists.
Sánchez [8] presented the approach which exploits the Web to extract ev-
idences that sustain a particular property of a given relationship. Specifically,
the axioms studied in this work are symmetrical, reflexive, functional, transitive,
inverse and inverse functional properties. They don’t extract property domains
and ranges and don’t extract using each property types mutually.
Building up Ontologies with Property Axioms from Wikipedia 3
Fleischhacker et al. [9] proposed a method of creating an RDF dataset with
property axioms. They extracted property subsumption, disjointness, transitiv-
ity, domain, range, symmetry, asymmetry, inverse, functionality, inverse func-
tionality, reflexivity and irreflexivity. They used DBpedia for evaluating their
approach. They don’t extract using each property types mutually.
3 Property Definition for Our Wikipedia Ontology
3.1 Extraction of Property Names
We extract properties by using the following two methods.
Extraction by Scraping Infobox This method has been proposed in the
past [3]. We extract a three-piece set of article-item-value in the Infobox as
“instance-property-property value”. In case the triple set is extracted directly
from the dump data, the meaning of these properties is likely to appear chal-
lenging to understand at a glance, or the properties may not be integrated.
Therefore, we convert properties in Infobox into HTML by Java Wikipedia API
(Bliki engine)1 here. Hereby, we can understand the properties’ meaning at a
glance and integrate properties.
Extraction by Scraping List Structure A large number of Wikipedia ar-
ticles have list structures. This method extracts triples from list structures as
“article name - heading name - each value”. At that time, we examine the cate-
gories that each article belongs to and collect a number of heading names that
appear in each category. This enables the extraction of the category that the ar-
ticle belongs to as a property domain. The procedure of this method is illustrated
by steps 1 - 4 below.
1. Extract categories and heading name from each article of Wikipedia dump
data.
2. Check occurrence rate of heading name from each category using (1).
3. Remove heading names that have a low occurrence rate from (2).
4. Extract heading name as property and each value of list structures as value
of the property from each article.
3.2 Extraction of Property Domain
A subject of the triple discussed in 3.1 is an article name as an instance. There-
fore, examining a category to which an article that is a subject belongs enables
the definition a domain of a property. Therefore, the Infobox template name is
extracted as a domain of each property in Infobox. Furthermore, we extract the
Wikipedia categories to which the article using the Infobox template belongs, as
1
http://code.google.com/p/gwtwiki
4 T. Kawakami et al.
a domain. Moreover, for articles for which Infobox is not defined, a hyper concept
that is extracted in “Extracting hypernymy-hyponymy from textual definition”
in [3] is extracted as a domain. Collating the domain that is extracted so far
with the class hierarchies extracted in [3], we extract the domain (considered as
classes) included in the class hierarchy. As a result, we can eliminate domains
that are not appropriate as classes. Finally, we remove a few occurrence rate
domain (five or less), and we extract the remaining as a domain.
3.3 Extraction of Property Range
It is comparatively more convenient to define a property domain because an
instance name as a subject of a property corresponds to an article name in a
property triple, and the Infobox template name of an article can be considered
to be a property domain. However, with regard to property ranges, an instance
name as a object of a property cannot conclude to an article name; therefore, it
is challenging to define the property ranges of all properties similar to property
domain. Thus, for property ranges, we use following two methods:
1. Extraction using class-instance relations
2. Extraction using is-a relations
They are generally linked if a word has already had an article in Wikipedia
and an article name corresponds to an instance name. Therefore, we match an
object of a triple (instance) and instances of class-instance relations. Then, we
extract classes as property ranges that the instance belongs.
Next, in order to extract property ranges that cannot be extracted by a
previous technique, we match the categories that an article (which has a name
identical to that of an object of a triple) belongs to and the classes of is-a relations
in [3]. Then, we extract the classes as property ranges.
3.4 Extraction of Property Hypernymy-Hyponymy Relations
We attempt extracting property hypernymy-hyponymy relations. We use the
following two methods.
Extracting by String Matching We extract the propety hypernymy-hyponymy
relations by using “backward string matching” in [3]. The procedure of this
method is shown following 1 - 4.
1. When a property A is set as “any string + property B”，we extract them
as property B rdfs:subPropertyOf property A.
2. In case they are set as “property A = any string + preposition + any string
+ property B”, we eliminate them.
3. The domains of properties A and B, the ranges of properties A and B must
be identical to or approximate the class hierarchy. Using this, we filter them.
4. We extract the remaining as property hypernymy-hyponymy relations.
Building up Ontologies with Property Axioms from Wikipedia 5
Figure 1 shows an example. In the figure, the properties “Partner” and “Do-
mestic partner” are matched, and their domains and ranges are identical. There-
fore, we extract them as property hypernymy-hyponymy relations.
rdfs:domain rdfs:range
Partner
People rdfs:subPropertyOf People
Domestic
partner
rdfs:domain rdfs:range
Fig. 1. Example of extraction of property hypernymy-hyponymy by string match
Extraction Using Triple from List Structures We match values of the
extracted propety names by scraping the Infobox triple and the list structure
in Wikipedia articles in each subject of the triples (instances). Then, when at
least one matched value exists, we extract the property name by scraping the
list structure and the property name by scraping the Infobox. Next, we compare
the number of instances of the propeties; the larger one is extracted as the
hypernymy, whereas, the smaller one is extracted as the hyponymy. Finally,
we match the domains and ranges in these candidate properties. Then, when
at least one matched properties exists, we extract these propeties as property
hyperhymy-hyporymy relations.
Figure 2 shows an example. In this case, a subject of the triple “Nevada
County, California” has a property “Communities” (which was extracted by
scraping a list structure) and its values “Nevada City” and a property “Largest
city” (which was extracted by scraping Infobox) and its values “Nevada City”.
Thus, we obtain “Community” as an upper property candidate and “Largest
city” as a lower property candidate. Subsequently, when we match the domains
and ranges of these properties, we know that these properties have identical
domain and same range. Thus, we extract “Community - Largest city” as a
property hypernymy-hyponymy relation.
3.5 Estimation of Property Types
Extracted properties by scraping Infobox have already classified Object type and
Data type. In this method, adding to Object type and Data type, we attempt to
estimate symmetric, transitive, functional, and inverse functional property types
using the triples whose extraction is described in Section 3.1.
Extraction of Symmetric Property First, we attempt to estimate the sym-
metric propety. We consider a subject X and a value Y of each property P ;
6 T. Kawakami et al.
Communities
Nevada County, Municipa
lities rdfs:subPropertyOf City Nevada City
California
rdfs:domain rdfs:range
Instance Instance
Largest city
Fig. 2. Example of extraction of property hypernymy-hyponymy using triple from list
structures
if a triple ”Y − P − X ” of P exist, we extract the propety P as a symmetric
property candidate α. In addition, we attempt to estimate the ratio of all the
triples of P to the triples extracted triples as symmetric relationship candidates.
Extraction of Transitive Property Next, we attempt to estimate transitive
property. We consider a subject X , a value Y and a value Z of a subject Y of
each propety P ; if a triple ”X − P − Z ” of P exists, we extract the propety P as
a transitive propety candidate. In addition, we attempt to estimate the ratio of
all the triples of P to the triples extracted as transitive relationship candidates
α.
Extraction of Functional Property Next, we attempt to estimate the func-
tional property. We condider a subject X , a value Y of each property P ; if P
has a value Y for each instance X , we extract the property P as a functional
property.
Extraction of Inverse Functional Property Moreover, we attempt to es-
timate the inverse functional property. We consider a subject X , a value Y of
each property P ; if P has only one instance X for each value Y , we extract the
property P as the inverse functional property.
Extracting Using Property Hypernymy-Hyponymy Relations In addi-
tion to extraction by the above four methods, we estimate property types by
using property hypernymy-hyponymy relations. We add property types to the
subproperty of the property that have types.
Building up Ontologies with Property Axioms from Wikipedia 7
3.6 Refining Extracted Relations
In this section, we propose refining methods of extracted property domain, prop-
erty range, symmetric property, transitive property, functional property, and in-
verse functional property. Here, we attempt to improve the accuracy by mutually
using each relation extracted by each method.
Refinement of Property Domain and Range In the method of Sections
3.2 and 3.3, multiple domains and ranges are linked to a property. In that case,
the property “BornPlace” links to the property domain “People” and “Japanese
people”. “Japanese people” is a subclass of “People” and can be regarded as
a redundant domain. Therefore, we consider removing such redundant domains
and ranges. Hereby, using the is-a relation already extracted, if the hyper class
of the extracted domain is also extracted as a domain and the latter has a high
inclusion rate, the former is removed.
Refinement of Symmetric Property and Transitive Property We refine
the symmetric and transitive proeprties extracted earlier. In the symmetric and
transitive properties, the domain and range of the property must match. There-
fore, we can filter them by using property domain and property range extracted
earlier.
4 Experimental Results and Observations
Now, we discuss the results and observations extracted from the Wikipedia dump
data as a source as of January 2017.
4.1 Results and Observations of Property Names
Results and Observations by Scraping Infobox From the Wikipedia dump
data, we extracted 8,311,427 infoboxes, 12,309 infobox templates, and 22,767,071
triples. The number of property names was 12,088.
According to the property type estimation, approximately 60% of the prop-
erties were classified as either owl:ObjectProperty or owl:DatatypeProperty. The
reason 40% of the properties were classified as Unknown was that the rule of
classification was insufficient. In the future, we need to take into consideration
that property values such as “small” can be literal. we extracted 1,000 samples
from all triples to estimate the precision. The precision was 93.1 ± 1.57%.
Hereby, We used the following expression (1) as a formula for 95% confidence
interval. In formula (1), n represents the number of samples, N represents pop-
ulation, and p̂ represents the estimated amount which is the number of accuracy
samples divided by the total number of samples.
√ √
n p̂(1 − p̂) n p̂(1 − p̂)
[p̂ − 1.96 (1 − ) , p̂ + 1.96 (1 − ) ] (1)
N n−1 N n−1
8 T. Kawakami et al.
Results and Observations by Scraping List Structure From the Wikipedia
dump data, we extracted 7,410,819 triples by scraping the list structure. The
number of property names was 2394. We can extract properties that cannot be
extracted by scraping Infobox, and they have numerous triples.
We extracted 1000 samples from all the triples to estimate the precision. The
precision was 81.6 ± 2.40%. A number of the errors were scraping errors, and
they were extracted when a large quantity of information was described in each
line of the list structures. For example, the property “Track listings” can be
observed in numerous Wikipedia albums or single articles of singers; however,
the list structures in these articles has generally described information such
as Writers, Released Date, and Length, apart from the Track listings. Thus,
Writers, Released Date and Length was extracted as values of the proeprty
“Track listings”. We consider it feasible to remove these errors to add more
detailed rules of list structures with each category that the articles belong to.
4.2 Results and Observations of Property Domain
We were able to extract 500,967 property domains for 12,088 extracted prop-
erties. The properties extracted by scraping Infobox has the Infobox template
as domains. Thus, we could define the property domain of all properties. In
addition, we define the property domain of 100 properties by scraping the list
structure in the Wikipedia articles. The property domains of 12,188 properties
were defined. Thus, 84.2% properties have property domains.
We used the property domain defined in DBpedia Ontology for the evalua-
tion. DBpedia Ontology defines properties, classes and property domains man-
ually. Therefore, there are numerous upper classes. However, our ontology uses
the Wikipedia category etc. Therefore, there are numerous lower classes. Conse-
quently, because it does not match in a straightforward comparison, we assumed
it to be correct to considered that the extracted property domain is included in
the property domain of DBpedia Ontology. In addition, the notation of prop-
erties is different from that of DBpedia’s properties in certain cases. By using
the Infobox template, we compare them with the Infobox items before being
mapped to the DBpedia ontology, to achieve consistency.
The precision was 94.9 ± 1.93%. Table 1 presents examples of property do-
mains. When a property has plurals property domains, we should integrate the
common upper concept classes. For example, the property “Staff” has “TV pro-
gram” and “TV drama” as property domain, and it has a number of classes as
property domains apart from these (e.g. “Radio program”, “Baseball team”).
Therefore, we attempt to integrate a common upper concept class in Section
3.6.
4.3 Results and Observations of Property Range
We were able to 78,117 propety ranges for 12,088 extracted properties. Similar
to the property domains, we evaluated property ranges using DBpedia Ontology.
The precision was 76.9 ± 3.68%. Table 2 presents examples of property range.
Building up Ontologies with Property Axioms from Wikipedia 9
Table 1. Example of extracted property domains
Property Property domain #Triples
Length Song 164,684
Nationality Person 167,429
Spouse Person 78,635
Table 2. Example of extracted property ranges
Property Property range #Triples
Starring Actor 401,656
Place of birth Town 890,271
Written by Person 158,123
For example of errors, “Native dress” as property range of the property “na-
tionality” was extracted. This is an error owing to an incorrect class-instance
relation. Numerous errors by this method are because of this reason. There-
fore, we consider that the precision of this method increases by improving the
precision of the class-instance relations. Furthermore, “Article on mathemat-
ics” was extracted as a range of the property “victory Frequency” or “num-
berOfAccommodations”. This is an error owing to an incorrect classificaion of
owl:ObjectProperty or owl:DatatypeProperty. Generally, the property type be-
comes owl:DatatypeProperty. Thus, the range becomes literal (rdfs:Literal).
4.4 Results and Observations of Property Hypernymy-Hyponymy
Relations
Results and Observations by String Matching We extracted 1297 property
hypernymy-hyponymy relations. We extracted 500 samples from all the relations
to estimate the precision. Thus, the precision was 89.0 ± 2.36%. As examples of
errors, “Movement - Tradition or movement” and “Period - Time period” are
extracted. We can prevent the former error “Movement - Tradition or movement”
by adding rules while removing unnecessary relations after string matching. The
latter is an example in which the properties with identical meaning are extracted
as hypernymy and hyponymy. In this case, it is feasible to assess whether they
have identical meaning by using external resources.
Results and Observations by Using Triple from List Structure We
were able to 242 property hyperhymy-hyponymy relations by using a triple from
the list structure. We determined all 242 relations’ truthfulness and falseness
manually. The precision was 19.4%. As an example of errors, numerous upper
properties and lower properties had identical meanings. Furthermore, numer-
ous upper properties and lower properties were reversed. Because the triples
extracted from Infobox are exceedingly large (although we had compared the
properties by the total number of property values), the property extracted from
10 T. Kawakami et al.
Infobox is an upper property, whereas the property extracted from the heading
is a lower property.
4.5 Results and Observations of Property Types
We attempted to estimate by using the methods described in Section 3.2 with
12,088 extracted properties and 22,767,071 extracted triples.
First, we attempted to estimate symmetric property. We were able to 13,371
symmetric relation triples and 89 symmetric property candidates. We determined
all 89 properties’ truthfulness and falseness manually. Thus, the precision was
51.7%. Moreover, when we regard the inclusion rate as the threshold value, the
precision is 92.3% (threshold 3%).
There are a number of properties by this method, including words such as
“Related” (“Related case”, “Related index”). However, this method could ex-
tract words such as “Former partner” and “Sister station” too.
Secondly, we attempted to estimate the transitive property. We were able
to 48,035 transitive relation triples and 165 transitive properties candidates.
Hereby, we excluded those which can be regarded as symmetric among the ex-
tracted relations because they are errors. As a results, We were able to 141
transitive property candidates excluding symmetric properties. We determined
all the 141 properties’ truthfulness and falseness manually. Thus, the precision
was 19.6%.
We consider the covering problem in Wikipedia to be a reason for such a
low accuracy. When we extract by this method, it is necessary to extract at
least three triples that become the transitive relations. Thus, it is necessary
to cover the information by Infobox or list structure as an identical property
name in the Wikipedia articles; however, such covered information is negligible.
For extracting transitive properties with high accuracy, we have to refine the
property names, integrate the same one, and attempt to extract another method
that use a part of the nostructuredness information in articles.
Thirdly, we attempted to estimate the functional property. We were able to
313,935 functional relation triples and 2,026 functional property candidates. We
extracted 500 samples from all the triples to estimate the precision. The precision
was 64.0 ± 3.59%.
The largest errors are that the number of triples is small and properties
that happen to have uniquely determined values of were extracted. Moreover,
because the information in Infobox is incomplete, we extracted it erroneously as
a functional property. We consider that errors owing to insufficient information
can be prevented by using information such as Wikipedia texts.
Fourthly, we attempted to estimate inverse functional property. We were able
to 10,187 inverse functional relation triples and 526 inverse functional property
candidates. We extracted 200 samples from all the triples to estimate the preci-
sion. The precision was 24.0 ± 4.03%.
There are problems of mark difference or synonym of property name, such
as “Main work” and the property “Notable work.” Therefore, we have to refine
the property names and integrate the same one to extract inverse functional
Building up Ontologies with Property Axioms from Wikipedia 11
property with high precision. Moreover, there were numerous errors owing to
mistakes caused by scraping of Infobox. Therefore, it is necessary to refine the
triple extraction method to improve accuracy.
Finally, we attempted extraction by using property hypernymy-hyponymy
relations. We were able to 4 symmetric properties, 129 transitive properties, 195
functional properties, and 108 inverse functional properties. We dertermined all
properties’ truthfulness and falseness manually. The precision was 75%, 7.8%,
68.7%, and 21.3% respectively. In order to increase the accuracy and extract nu-
merous relations, it is necessary to reconsider the estimation method of property
type and the extraction method of subproperties.
4.6 Results and Observations of Refining Extracted Relations
Results and Observations of Refining of Property Domain and Range
We refined extracted domains and ranges. As a result, the average and variance
of the domains and ranges per property changed, as shown in the Table 3. Even
after refinement, the number of domains and ranges per property is large, and
there are variations depending on the properties. This is because the hierarchical
structure of the class hierarchy is incomplete, and it is necessary to refine the
class hierarchy in the future.
Table 3. Comparison of average and variance before and after refining
Before refinement After refinement
Average Variance Average Variance
rdfs:domain 75.9 329094.1 38.5 64317.6
rdfs:range 33.7 39786.9 26.8 21672.6
Results and Observations of Refinement of Symmetric Property and
Transitive Property We refined the extracted symmetric properties and tran-
sitive properties. As a result, the number of extracted relations and accuracy
changed, as illustrated in the Table 4. Although the precision marginally in-
creased with respect to the transitive properties, the precision marginally de-
creased with respect to the symmetric properties. The reasons for this are in-
sufficient refinement and insufficient extraction of domains and ranges. In the
future, it is necessary to unify classes and improve the refinement methods of
domains and ranges.
5 Conclusion
In this paper, we proposed and evaluated a method of constructing a large-
scale and general-purpose Ontology with property axioms using Wikipedia as
12 T. Kawakami et al.
Table 4. Comparison of number of extracted relations and precision
Before refinement After refinement
Number of relations Precision Number of relations Precision
Symmetric 89 51.7% 47 51.5%
Transitive 141 19.6% 77 24.7%
the resource. Through this study, we demonstrated that Wikipedia is a valuable
resource for ontology learning for property axioms. It is feasible to extract prop-
erty domain, property range, and property hypernymy-hyponymy relations etc.
from Wikipedia. A few of the proposed methods can also be applied to DBpe-
dia. These can be considered effective for constructing costless and large-scale
ontologies in ontology learning.
In the future, we plan to improve the precision of each method．Moreover,
we plan to integrate properties and unify classes extracted in [3] by using a upper
ontology such as WordNet. Also, we intend to provide our Wikipedia Ontology.
References
1. Miller, G.A.: WordNet: A Lexical Database for English. Communications of the
ACM Vol. 38, No. 11: 39-41. (1995)
2. Lenat, D.B. and Guha, R.V.: Building Large Knowledge Based Systems. Addison-
Wesley (1990)
3. Kawakami, T., Morita, T., Yamaguchi, T.: Building Wikipedia Ontology with More
Semi-structured Information Resources. JIST2017, LNCS, vol.10675, pp.3-18
4. Auer, S., Bizer, C., Kobilarov, G., Lehmann, J., Cyganiak, R., Ives, Z.: DBpedia:
A Nucleus for a Web of Open Data. Lecture Notes in Computer Science, Springer
Berlin / Heidelberg, pp.722-735(2007)
5. Suchanek, F.M., Kasneci, G., Weikum, G.: YAGO: A Large Ontology from
Wikipedia and WordNet. Elsevier Journal of Web Semantics (2008)
6. Hoffart, J., Suchanek, F., Berberich, K., Weikum, G.: YAGO2: A Spatially and
Temporally Enhanced Knowledge Base from Wikipedia. Reserch Report MPI-I-
2010-5007，Max-Planck-Institut fur Informatik (2010)
7. Mahdisoltani, F., Biega, J., Suchanek, F.M.: A Knowledge Base from Multilingual
Wikipedias. In: CIDR (2015)
8. Sánchez, D., Moreno, A., Del Vasto-Terrientes, L. (2012): Learning relation axioms-
from text: An automatic web-based approach. Expert Systems with Applications,
39, 5792 ‒ 5805.
9. Fleischhacker, D., Völker, J., Stuckenschmidt, H.: Mining rdf data for property
axioms. In: Meersman, R., Panetto, H., Dillon, T., Rinderle-Ma, S., Dadam, P.,
Zhou, X., Pearson, S., Ferscha, A., Bergamaschi, S., Cruz, I.F. (eds.) OTM 2012,
Part II. LNCS, vol. 7566, pp. 718 ‒ 735. Springer, Heidelberg (2012)