=Paper=
{{Paper
|id=None
|storemode=property
|title=Argumentation-based Inconsistencies Detection for Question-Answering over DBpedia
|pdfUrl=https://ceur-ws.org/Vol-1064/Cabrio_Argumentation.pdf
|volume=Vol-1064
|dblpUrl=https://dblp.org/rec/conf/semweb/CabrioCVG13a
}}
==Argumentation-based Inconsistencies Detection for Question-Answering over DBpedia==
https://ceur-ws.org/Vol-1064/Cabrio_Argumentation.pdf
Argumentation-based Inconsistencies Detection
for Question-Answering over DBpedia
Elena Cabrio, Julien Cojan, Serena Villata, and Fabien Gandon
INRIA Sophia Antipolis, France
{firstname.lastname}@inria.fr
Abstract. In the Web of Data, the pieces of information obtained by
querying distributed SPARQL endpoints may provide different results for
the same query. Moreover, the combination of these query results may
lead to an inconsistent set of information about the same topic. In par-
ticular, the problem of reconciling information obtained by distributed
SPARQL endpoints is encountered in question-answering systems over
linked data, where different SPARQL endpoints are queried to retrieve
the answer to the user’s question. In this paper, we propose to address
this problem by adopting argumentation theory to reason over inconsis-
tent information sets, and provide nevertheless a unique and motivated
answer to the user. We implement and evaluate our approach on QAKiS
(Question Answering WikiFramework-based system), that exploits mul-
tilingual chapters of DBpedia as RDF data sets to be queried using a
natural language interface.
1 Introduction
In the Web of Data, the combination of the information items concerning a
single real-world object coming from different data sources, e.g., the results of
a single SPARQL query on different endpoints, may lead to an inconsistent
results set. This is an open problem for consuming in the Web of Data since
these inconsistencies mine the overall quality of the data itself. In particular,
this problem arises while querying the multilingual chapters of DBpedia [12].
Such chapters, well connected through Wikipedia instance interlinking, can in
fact contain different information with respect to the English version. Assuming
we wish to query a set of multilingual DBpedia SPARQL endpoints with the
same query, the answers we collect can be either identical, or one can subsume
the other, or they can be contradictory.
In this paper, we answer the following research question: How to reconcile
information provided by the multilingual chapters of DBpedia to obtain a con-
sistent results set? This issue is particularly relevant to Question Answering
(QA) systems over DBpedia [11], where the user expects a unique (and possi-
bly correct) answer to her factual natural language question. In this scenario,
another open issue is to motivate the answer the system provides to the user in
such a way that the overall question answering system appears as transparent
to her, and, as a consequence, more reliable. Following these considerations, our
research question breaks down into the following subquestions:
1. How to semantically relate information items provided by different multilin-
gual chapters of DBpedia to detect possible inconsistencies?
2. How to compute the acceptability degree of information items to provide a
unique answer?
3. How to motivate and explain the reasons behind the answer provided by the
question answering system?
First, we need to detect the semantic relations which relate each piece of infor-
mation to the others returned by the different sources, i.e., SPARQL endpoints.
In particular, two kinds of relations can be highlighted: a conflict relation such
that the information items represent contradicting answers to a single question
(e.g., given the question “Which is the capital of Italy? ”, the answers “Paris”
and “Rome” conflict), and a support relation such that the information items are
semantically connected in a way that one item may be derived from the other
by means of an ontology (e.g., given the question“In which place was William
Shakespeare born? ”, the answers “Stratford-upon-Avon” and “England ” do not
actually conflict because Stratford-upon-Avon is located in England).
Second, we adopt abstract argumentation theory [7] to reason over the incon-
sistencies among a set of information items called arguments, and to return a
consistent (sub)set of them. Roughly, an abstract argumentation framework is a
directed labeled graph whose nodes are the arguments and the edges represent a
conflict relation. Since there are situations where the sources provide the same
answer or a subsumed one, we need to represent also a positive relation among
the arguments. For this reason, we rely on bipolar abstract argumentation [4]
where also a support relation is considered. We compute the acceptability degree
of the arguments depending on the confidence assigned to their sources [5].
Third, the overall argumentation framework together with the acceptability
degree of each argument is used to motivate to the user the answer the system
returns. We evaluate our approach through its integration in QAKiS [3], that
exploits DBpedia multilingual chapters as RDF data sets to be queried using a
natural language interface. The argumentation module is embedded to provide
a (possibly unique) answer whose acceptability degree is over a given threshold,
and the graph structure linking the different answers is provided as motivation.
In this paper, we do not address the issue of linked data quality assessment and
fusion (Sieve [13]), and we do not improve DBpedias instances alignment. We do
not use argumentation theory to find agreements over ontology alignments [14].
The remainder of the paper is as follows: Section 2 provides the basic notions
of argumentation theory. Section 3 presents our argumentation-based framework
for inconsistencies detection. In Section 4 our approach is evaluated on QAKiS.
Section 5 compares the existing research with the proposed approach.
2 Background
A Dung-style abstract argumentation framework [7] (AF) aims at represent-
ing conflicts among elements called arguments through a binary attack (i.e.,
conflict) relation. The need to introduce also a positive relation among the ar-
guments, i.e., a support relation, leads to the proposal of so called bipolar argu-
mentation frameworks [4] (BAF). An example of BAF is visualized in Figure 1.b
where the dotted edge represents the support relation. This kind of framework
allows to reason about these conflicts to detect, starting by a set of arguments
and the conflicts among them, which are the accepted arguments. Accepted ar-
guments are those arguments which are considered as believable by an external
evaluator, who has a full knowledge of the argumentation framework. Roughly,
an argument is accepted (i.e., labelled in) if all the arguments attacking it are
rejected, and it is rejected (i.e., labelled out) if it has at least an argument attack-
ing it which is accepted. Figure 1.a shows an example of abstract argumentation
framework. The arguments are visualized as circles, and the attack relation is
visualized as edges. Gray arguments are the accepted ones. We have that argu-
ment a attacks argument b and b attacks a, and argument c attacks a. Using
Dung’s acceptability semantics [7], the set of accepted arguments is {b, c}.
However, associating a crisp label, i.e., in or out, to the arguments is lim-
iting in a number of real life situations where a numerical value expressing
the acceptability degree of each argument is required [8, 5, 9]. In particular, da
Costa Pereira et al. [5] propose a fuzzy labeling algorithm to account for the
fact that arguments may originate from sources that are trusted only to a cer-
tain degree. They define α a fuzzy labeling iff for all arguments A, α(A) =
min{A(A), 1 − maxB:B→A α(B)} where A(A) is given by the trust degree of the
most reliable source that offers argument A. Consider the example in Figure 1.a,
if we have A(a) = A(b) = A(c) = 0.8, then the algorithm returns the following
labeling: α(a) = 0.2 and α(c) = α(b) = 0.8.
3 The framework
This section describes our
framework to detect the in-
consistencies among the re-
c a b c a b
sults obtained querying the
SPARQL endpoints of multi-
lingual DBpedia chapters: i) (a) d (b)
how we assign an attack or a
support relation between two
Fig. 1: Example of (a) AF , (b) BAF .
pieces of information depend-
ing on the semantic relation
linking them (Section 3.1),
and ii) how compute a fuzzy evaluation of the arguments (Section 3.2).
3.1 From ontological relations to argumentation relations
Given a set of answers to a certain factual query, each provided by a different
endpoint, we define and apply our algorithm to assign the relations of support
and attack between two arguments (where an argument is an answer indepen-
dently provided by a specific SPARQL endpoint). Multilingual chapters of DB-
pedia [12], well connected through Wikipedia instance interlinking, can in fact
contain different information with respect to the English version. Assuming we
wish to query a set of multilingual DBpedia SPARQL endpoints with the same
query, the answers we collect can be either identical, or one can subsume the
other, or they can be contradictory. Each endpoint is assigned a confidence score,
according to the probability that such data set is reliable with respect to the
information items it contains. In this case study, we assign an apriori confidence
score to the endpoints according to their dimensions and solidity in terms of
maintenance, but other methods to assign such scores can be explored as future
work (e.g., letting the user to select the endpoints that she considers as more
reliable).
Fig. 2: Examples of support relations: identity (a,b), subsumption (c,d).
Given a certain query, if two endpoints provide identical answers, we identify
a positive relation between them (that we call identity) which is then translated
into a support relation between such arguments. For instance, in Figure 2a both
the French and the English DBpedia SPARQL endpoints provide Italy as answer
to Where is the Colosseum located?. The algorithm assigns therefore a support
relation between such arguments (double arrows in the figures), and merges them
into a unique argument with value Italy. We do not consider these arguments
as independent for two reasons: first, every other answer conflicting with one of
them is also conflicting with all the others, and second, given that this argument
is shared among several sources then it is highly acceptable, i.e., reliable. The
confidence score of this new argument is calculated as the arctangent of the
confidence scores of the endpoints providing such answer (max value = 1). We
benefit from the sameAs links between the translation of the same word in
DBpedia multilingual chapters, and we consider also the case reported in Figure
2b as identity, since both answers contain the same value expressed in different
languages (i.e., Italia in Italian, and Italy in English).
Figure 2 (c,d) reports another positive relation between arguments, i.e., what
we call subsumption. This case arises when one of the obtained answers is more
specific than the other, both in terms of i) spacial/geographical relation, e.g.,
x is located in y (Figure 2d: Rome is located in Italy); ii) hyperonymy (is a
relation) as in Figure 2c (Gibson J-160E is a Guitar : this example shows two
possible answers to the question Which instruments did John Lennon play? ).
Our algorithm exploits external sources of semantic knowledge to detect the
relations between the arguments (e.g., GeoNames1 for geographical entities, and
the DBpediaYAGO class hierarchy2 [15]).
Once such relations are found and verified, the algorithm sets a support re-
lation between the two arguments following the direction from the most specific
argument toward the more general one (double arrows in Figure 2).
Fig. 3: Examples of attack relation: a) direct attack, b) additional attack.
In case the answers provided by two different endpoints for a certain query
are different, and no subsumption relation between them is identified, then the
algorithm assigns an attack relation between such arguments. For instance, an
attack arises between the arguments visualized in Figure 3a, since the answers
provided by the Italian and by the English DBpedia endpoints to the query
concerning the location of the Colosseum are contradictory. It should be noted
that the attack relation is always bidirectional in our framework (i.e., the two
arguments conflict with each other, as shown by simple arrows in Figure 3).
A subtler case of attack is reported in Figure 3b. Here we have a support rela-
tion between the arguments Roma and Italy provided respectively by the Italian
and the English endpoints (subsumption relation, as in the example visualized in
Figure 2). Then, a third endpoint, i.e., the French one, provides the value France
as answer to the same query. Our algorithm assigns an attack relation between
this argument and the argument Italy, and it automatically adds an additional
attack from France to the argument Roma which is supporting the argument it
attacks (i.e., Italy). This additional attack results from the assumption that if
an argument attacks Italy, it attacks also Rome since Rome is in Italy.
The last case we consider in our algorithm is when each endpoint provides a
list of values (i.e., a list) as answer to a certain query (e.g., non-functional prop-
erties in DBpedia). For instance, in the example reported in Figure 4 the English
DBpedia endpoint provides both Guitar and Voice as instruments played by J.
Lennon, while the French endpoint provides both Guitare and Harmonica. In
this case, our algorithm assigns a support relations between Guitar and Guitare,
and it merges them in a unique value (see identity in Figure 2). Moreover, dif-
1
http://www.geonames.org/
2
http://dbpedia.org/Downloads38#download-links-to-yago2
Fig. 4: Example of lists answers.
ferently from the case illustrated in Figure 3, in case of lists we do not consider
arguments of the same list as conflictual. For this reason, we limit the assignment
of the attack relation to those arguments which are not provided together in the
same list. As illustrated in Figure 4, no attack relation is assigned between, e.g.,
Harmonica and Guitar, since both answers are provided by the French endpoint.
In this section we have described the algorithm that, given the answers pro-
vided by the endpoints to a single query, highlights the support/attack relation
among them. Next section explains how such list of arguments and relationships
is sent to the argumentation module that calculates the arguments’ acceptability
degree (i.e., the arguments that will be proposed to the user as more reliable).
3.2 Arguments evaluation and answer’s motivation
Since we need to take into account the confidence associated to an information
source when computing the set of accepted arguments and their own acceptabil-
ity degree, we go beyond standard argument semantics introduced by Dung [7],
and we rely on the computation of fuzzy confidence-based degrees of accept-
ability. In particular, the fuzzy labeling algorithm proposed by da Costa Pereira
et al. [5] exploits a scenario where the arguments cannot be evaluated in the
same way because of the confidence assigned to their source. In order to account
for this fact and to consider also a positive, i.e., support, relation among the
arguments, in addition to the attack relation used in [5] for the computation
of the fuzzy labels of the arguments, in this paper we propose a bipolar fuzzy
labeling algorithm. Let A be a fuzzy set of trustful arguments, and A(A) be the
membership degree of argument A in A, we have that A(A) is given by the trust
degree of the most reliable (i.e., trusted) source that offers argument A, and it is
defined as follows: A(A) = maxs∈src(A) τs where τs is the degree to which source
s ∈ src(A) is evaluated as reliable [5]. We follow da Costa Pereira et al. [5] in
this choice, which implies an “optimistic” assignment of the labels. In case a
pessimistic assignment is preferred, the min operator has to be used.
We now extend the definition of fuzzy labeling for standard abstract argu-
mentation into a bipolar fuzzy labeling. Note that in this paper we assume that
the following two constraints hold: an argument cannot attack and support at
the same time another argument, and an argument cannot support an argument
attacking it, and vice versa. These constrains are not on the computation of the
acceptability degree of the arguments, but they underlie the construction of the
argumentation framework itself.
Definition 1. Let hA, →, ⇒i be an abstract bipolar argumentation framework
where A is a fuzzy set of (trustful) arguments, →⊆ A × A and ⇒⊆ A × A
are two binary relations called attack and support, respectively. A bipolar fuzzy
labeling is a total function α : A → [0, 1].
Such an α may also be regarded as (the membership function of) the fuzzy set
of acceptable arguments where the label α(A) = 0 means that the argument is
outright unacceptable, and α(A) = 1 means the argument is fully acceptable. All
cases inbetween provide the degree of the acceptability of the arguments which
may be considered accepted at the end, if they overcome a given threshold. In [5],
the acceptability of an argument cannot be greater than the degree to which
the arguments attacking it are unacceptable: α(A) ≤ 1 − maxB:B→A α(B). This
constraint is reformulated in bipolar fuzzy labeling as follows: the acceptability of
an argument cannot be greater than the degree to which the arguments attacking
it are unacceptable unless there exists at least one argument supporting it.
The rationale behind this constraint is the following: if the set of supports is
empty, then the standard fuzzy labeling is used; otherwise, the support of an ar-
gument towards another is intended to augment the acceptability degree of such
supported argument. As we will show by means of examples, the acceptability
degree of an argument which is both supported and attacked depends on the
confidence assigned to the sources proposing the supporter and the attacker of
such argument. Using the above constraint, we obtain the following definition
Definition 2. (Bipolar Fuzzy Labeling) Let α be a bipolar fuzzy labeling. We say
that α is a bipolar fuzzy labeling iff, for all arguments A,
α(A) = avg{min{A(A), 1 − maxB:B→A α(B)}; maxC:C⇒A α(C)}.
da Costa Pereira et al. [5] show that the convergence speed of the labeling
algorithm is linear (as their proof of convergence suggests) since in practice a
small number of iterations is enough to get so close to the limit that the error is
less than the precision with which the membership degrees are represented in the
computer. The bipolar fuzzy labeling algorithm presented here is a variant of the
one developed in [5] thus on account of this fact, the computational feasibility
result can be imported here. Table 1 reports the iterations performed by the
algorithm to assign the labels to the arguments of a BAF .
When the argumentation module receives the couples of arguments linked by
the appropriate relation and the degree of confidence associated to each source,
the bipolar fuzzy labeling algorithm is raised on the argumentation framework
to obtain the acceptability degree of each argument. This step returns also the
overall bipolar argumentation framework where each argument is linked to its
source, and the acceptability degree is associated to the arguments. This overall
view is then used to explain to the user how the QA system comes to find this
answer, thus its motivations. This step is necessary to prevent the user from
seeing the QA system as a black box, but to understand the reasons behind
correct, and more importantly, erroneous answers.
The fact that an argumentation framework
can be used to provide explanations [1] is Table 1: BAF : A → B, B →
one of the reasons behind the choice of this C, C → A, D ⇒ C
formalism to detect inconsistencies. Other t αt (A) αt (B) αt (C) αt (D)
possible solutions to rank a set of informa- 0 1 0.4 0.2 1
tion items would be (possible weighted) vot- 1 0.9 0.2 0.6 1
ing mechanisms, where the preferences of 2 0.65 0.15 ↓ ↓
some voters, i.e., the most reliable informa- 3 0.52 0.25
tion sources, carry more weight than the pref- 4 0.46 0.36
erences of other voters. We choose to rely 5 0.43 0.4
on argumentation-based inconsistency detec- 6 0.41 ↓
tion instead of adopting a voting system to 7 0.4
rank the answers provided by the QA sys- 8 ↓
tem because the latter does not consider the
presence of semantic relations (of positive or
negative nature) among the items within the list. The additional value of an
argumentation-based approach is the graph-based visualization of the answers.
4 Framework integration in a QA system and evaluation
To evaluate the argumentation-based framework to detect inconsistencies that
we described in Section 3 in a real setting, we integrate it in an existing Question-
Answering system, QAKiS [3] (Section 4.1). QAKiS allows to query multilingual
chapters of DBpedia as RDF data sets using a natural language interface, and
its architecture can be flexibly modified to account for the proposed extension.
4.1 QA system
QAKiS3 (Question Answering wiKiFramework-based System) [3] addresses the
task of QA over structured knowledge-bases (e.g., DBpedia), where the relevant
information is expressed also in unstructured forms (e.g., Wikipedia pages). It
implements a relation-based match for question interpretation, to convert the
user question into a query language (e.g., SPARQL). More specifically, it makes
use of relational patterns (automatically extracted from Wikipedia and collected
in the WikiFramework repository [3]), that capture different ways to express a
certain relation in a given language. QAKiS is composed of four main modules
(Figure 5): i) the query generator takes the user question as input, generates
the typed questions, and then generates the SPARQL queries from the retrieved
patterns; ii) the Named Entity (NE) Recognizer; iii) the pattern matcher
takes as input a typed question, and retrieves the patterns (among those in the
repository) matching it with the highest similarity; and vi) the sparql package
handles the queries to send to multiple multilingual DBpedia endpoints [2]. The
actual version of QAKiS targets questions containing a Named Entity related
3
http://dbpedia.inria.fr/qakis/
to the answer through one property of the ontology, as Which river does the
Brooklyn Bridge cross?. Such questions match a single pattern, i.e., one relation.
Fig. 5: QAKiS workflow
We embed into the QAKiS architecture the argumentation module (Figure 5),
that collects the answers obtained from the different endpoints and applies the
algorithm described in Section 3.1 to identify the arguments and the relations
among them, and then it runs the bipolar fuzzy labelling algorithm (Section 3.2)
to detect inconsistencies and return the more reliable arguments.
4.2 Data set
To run our experiments, we extract from the reference data set of QALD-2
(Question Answering over Linked Data challenge)4 the questions that the current
version of QAKiS is built to address (i.e. questions containing a NE related to
the answer through one property of the ontology), corresponding to 26 questions
in the training and 32 in the test sets. The discarded questions require either
some forms of reasoning (e.g., counting or ordering) on data, aggregation (from
datasets different from DBpedia), involve n-relations, or are boolean questions.
We consider these 58 questions as the reference data set for our experiments.
4.3 Results and error analysis
To evaluate the validity of the proposed approach, we run the questions con-
tained into our reference datasets on the English, German and French chapters
of DBpedia (confidence scores: 0.7, 0.6 and 0.5 respectively). We chose these
three chapters based on i) their dimensions and the robustness of their end-
points, and ii) the presence of the SameAs relation among the translations of
the words in the different languages. Since the questions of QALD-2 dataset were
4
http://greententacle.techfak.uni-bielefeld.de/~cunger/qald/index.php?x=
challenge&q=2
created to query the English chapter of DBpedia only, it turned out that only
in 25 cases out of 58 at least two endpoints provide an answer (in all the other
cases the answer is provided by the English chapter only, not useful for our pur-
poses). For instance, given the question List the children of Margaret Thatcher
the English DBpedia provides Mark and Carol Thatcher as answers, while the
French one provides only the answer Mark Thatcher. Or given the question How
many employees does IBM have?, the English and the German DBpedia provide
426751 as answer, while the French DBpedia provides 433362 and 2010.
We evaluate our approach with two sets of experiments: in the first case, we
start from the answers provided by the different DBpedia endpoints to the 25
questions, and we run our argumentation-based algorithm on it. In the second
case, we add QAKiS in the loop, meaning that the data we use as input for
the argumentation module are directly produced by the system, as explained in
Section 4.1. In this second case, the input are the 25 natural language questions.
Table 2 reports the results we obtained for the two experiments over 24 questions,
one question timed out. We evaluated both the ability of the argumentation mod-
ule to correctly identify and (if necessary merge) the answers from the different
endpoints (1st row), and its ability to assign the correct relations among such ar-
guments (2nd row), w.r.t. a manually annotated goldstandard of arguments and
their relations (in total, over the 25 questions, 90 items should be recognized as
arguments-answers, and 219 attack and 3 support relations should be detected).
For instance, given the examples before, for the question How many employees
does IBM have? the algorithm generates 4 arguments (where the answer 433362
provided by both the English and German DBpedia is merged and its confidence
score augmented, see Section 3) and 12 relations (attacks among the different
values). Applying the fuzzy labeling algorithm, the answer provided with the
highest confidence is 433362 (in future work, we could allow some approxima-
tion, merging values ∼433000 to this value). For List the children of Margaret
Thatcher, the algorithm identifies two arguments (i.e. Mark and Carol ), and no
relations, since the DBpedia property child allows for list answers. Both answers
are provided to the user, where Mark has a highest confidence score.
Most of the errors in identifying the arguments are due to the missing SameAs
links in DBpedia: the algorithm is therefore not able to merge translations of
the same answer, considering them as different. Wrong relation assignments are
mainly due to missing attacks among arguments (in particular for numerical val-
ues, or for arguments from the same endpoint). Concerning the second column
of Table 2, since QAKiS performances are about ∼50%, the results are ob-
tained accordingly (the argumentation-based module is biased by QAKiS mis-
takes in submitting the query to DBpedia). The average computation cost of
the argumentation-based algorithm is high (∼124 seconds), mainly due to the
n-answers questions. Considering only 1-answer question, the computation cost
drops to 5 seconds. The complexity is quadratic, at least one SPARQL query is
sent for each couple of answers. This could be improved by importing data for
each answers individually and then process it locally. We are currently investi-
gating this solution to optimize the algorithm.
Task Argumentation module QAKiS+Argum. module
Precis. Recall F-meas. Precis. Recall F-meas.
Argument identification 0.95 1 0.97 0.71 0.75 0.73
Relation assignment 0.71 0.73 0.72 0.54 0.56 0.55
Table 2: Performances over 24 questions from QALD-2 test set
5 Related work
State of the art QA systems over Linked Data generally address the issue of
question interpretation mapping a natural language question to a triple-based
representation (see [11] for an overview). Moreover, they examine the potential
of open user friendly interfaces for the SW to support end users in reusing and
querying the SW content. None of these systems provides a mechanism to detect
the inconsistencies among the set of items composing the answers, and none of
them allows the user to understand the reasons behind the retrieved answer.
Several works address alignment agreement based on argumentation theory.
More precisely, Laera et al. [10] address alignment agreement relying on argu-
mentation to deal with the arguments which attack or support the candidate
correspondences among ontologies. Doran et al. [6] propose a methodology to
identify subparts of ontologies which are evaluated as sufficient for reaching an
agreement, before the argumentation step takes place, and dos Santos and Eu-
zenat [14] present a model for detecting inconsistencies in the selected sets of
correspondences relating ontologies. In particular, the model detects logical and
argumentation inconsistency to avoid inconsistencies in the agreed alignment.
The framework we propose has common points with this line of works, i.e., the
use of argumentation theory to select a consistent set of information items, but
the scenario in which the two approaches are exploited is different and this leads
to a different addressed issues and proposed solutions.
6 Conclusions
In this paper, we propose an automatic framework to detect the inconsistencies
which may arise into a set of answers provided by a QA system over linked
data. These inconsistencies are due to i) heterogeneous answers provided by the
same information source, e.g., a single SPARQL endpoint, or ii) heterogeneous
answers from distributed information sources. Our framework adopts bipolar
abstract argumentation to provide an overall view of the information items pro-
vided by the single sources. This argumentation framework is built thank to the
semantic relations among the information items automatically extracted using
ontological knowledge, e.g., DBpedia ontology. The framework is then evaluated
to assign the acceptability degree of the arguments using the bipolar fuzzy la-
beling algorithm. We have shown the feasibility of our approach on the QAKiS
question answering system on the multilingual chapters of DBpedia.
There are several points to be addressed as future work. First, we assign to
each information source a confidence degree which is derived from the coverage
of such DBpedia version. However, another possibility is to leave the user itself
to assign the confidence degree to the information sources. We plan to embed
this feature in the QAKiS interface. Second, we have to extend the set of on-
tologies we consider in order to be able to detect further relations (positive and
negative) among the information items. Finally, the user evaluation should not
be underestimated and we will perform an evaluation campaign to verify which
kind of visualization of the motivations is more usable by consumers.
References
1. Bench-Capon, T.J.M., Lowes, D., McEnery, A.M.: Argument-based explanation of
logic programs. Knowl.-Based Syst. 4(3), 177–183 (1991)
2. Cabrio, E., Cojan, J., , Gandon, F., Hallili, A.: Querying multilingual DBpedia
with QAKiS. In: Procs ESWC 2013 (demo paper) (2013)
3. Cabrio, E., Cojan, J., Aprosio, A.P., Magnini, B., Lavelli, A., Gandon, F.: Qakis:
an open domain qa system based on relational patterns. In: Procs of ISWC 2012
(Posters & Demos). vol. 914 (2012)
4. Cayrol, C., Lagasquie-Schiex, M.C.: Bipolarity in argumentation graphs: Towards
a better understanding. In: Procs of SUM 2011, LNCS, vol. 6929, pp. 137–148.
Springer (2011)
5. da Costa Pereira, C., Tettamanzi, A., Villata, S.: Changing one’s mind: Erase or
rewind? In: Procs of IJCAI 2011. pp. 164–171. IJCAI/AAAI (2011)
6. Doran, P., Tamma, V., Palmisano, I., Payne, T.: Efficient argumentation over on-
tology correspondences. In: Procs of AAMAS 2009. pp. 1241–1242 (2009)
7. Dung, P.: On the acceptability of arguments and its fundamental role in non-
monotonic reasoning, logic programming and n-person games. Artif. Intell. 77(2),
321–358 (1995)
8. Dunne, P.E., Hunter, A., McBurney, P., Parsons, S., Wooldridge, M.: Weighted ar-
gument systems: Basic definitions, algorithms, and complexity results. Artif. Intell.
175(2), 457–486 (2011)
9. Hunter, A.: A probabilistic approach to modelling uncertain logical arguments. Int.
J. Approx. Reasoning 54(1), 47–81 (2013)
10. Laera, L., Blacoe, I., Tamma, V., Payne, T., Euzenat, J., Bench-Capon, T.: Argu-
mentation over ontology correspondences in MAS. In: Procs of AAMAS 2007. pp.
1–8 (2007)
11. Lopez, V., Uren, V.S., Sabou, M., Motta, E.: Is question answering fit for the
semantic web?: A survey. Semantic Web 2(2), 125–155 (2011)
12. Mendes, P., Jakob, M., Bizer, C.: DBpedia: A multilingual cross-domain knowledge
base. In: Procs of LREC 2012. ELRA (2012)
13. Mendes, P.N., Mühleisen, H., Bizer, C.: Sieve: linked data quality assessment and
fusion. In: Procs of the Joint EDBT/ICDT Workshops. pp. 116–123. ACM (2012)
14. dos Santos, C.T., Euzenat, J.: Consistency-driven argumentation for alignment
agreement. In: Procs of OM 2010, CEUR Workshop Proceedings 689 (2010)
15. Suchanek, F.M., Kasneci, G., Weikum, G.: Yago: a core of semantic knowledge. In:
Procs of WWW 2007. pp. 697–706. ACM (2007)