=Paper=
{{Paper
|id=None
|storemode=property
|title=On combining numerical optimization techniques with a belief merging approach
|pdfUrl=https://ceur-ws.org/Vol-2264/paper5.pdf
|volume=Vol-2264
|authors=Oscar Chávez-Bosquez,Pilar Pozos-Parra,Betania Hernández-Ocaña
|dblpUrl=https://dblp.org/rec/conf/lanmr/Chavez-BosquezP18
}}
==On combining numerical optimization techniques with a belief merging approach==
https://ceur-ws.org/Vol-2264/paper5.pdf
On combining numerical optimization techniques
with a belief merging approach
Oscar Chávez-Bosquez1 , Pilar Pozos-Parra2 , and Betania Hernández-Ocaña1
1
División Académica de Informática y Sistemas, Universidad Juárez Autónoma de
Tabasco {oscar.chavez,betania.hernandez}@ujat.mx
2
Instituto Tecnológico Superior de Occidente mariapozos@iteso.mx
Abstract. Metaheuristics are well-known numerical algorithms success-
fully used to solve different combinatorial optimization problems. On
the other hand, belief merging is a logic-based technique used to fu-
sion potentially conflicting pieces of information coming from different
sources. Although these two techniques solve different kinds of problems,
there are many real-life scenarios requiring both numerical and symbol-
ical approaches. In this paper, we propose a general hybrid framework
combining metaheuristics and belief merging. At a very high level, our
proposal consists of two black-box modules: an optimization engine and
a belief merger. For the first module we propose implementing a Tabu
Search, a fast, robust and reliable metaheuristic. For the second module
we propose using the ∆ps (PS-Merge) belief merging operator, a versatile
operator whose main advantage is that it can operate over inconsistent
belief bases; it outperforms other operators in a previous benchmark
study. Finally, we describe our proposal and provide insights about its
usefulness solving a motivating example.
Keywords: Hybrid framework · Metaheuristics · Knowledge Represen-
tation.
1 Introduction
Combinatorial optimization seeks to obtain the “optimal” solution among a set of
possibilities. Examples range from popular games like Sudoku, classic problems
like the Boolean Satisfiability, to real-life scenarios like the Traveling Salesman
Problem. The solution is encoded with discrete variables and typically consist
of an integer number, a subset, a permutation, or a graph structure [1].
Definition 1. A Combinatorial Optimization Problem (COP) P = (S, f ) is
defined by:
– A set of variables X = {x1 , . . . , xn }.
– Variable domains D1 , . . . , Dn .
– Constraints among variables.
– An objective function f to be maximized (or minimized), where f : D1 ×
· · · × Dn → R+ .
51
– The set of all possible, feasible assignments is S = {s = {(x1 , υ1 ), . . . , (xn , υn )}
| υi ∈ Di , s satisfies all the constraints}
– S is the solution space, as each element of the set can be seen as a candidate
solution. To solve a COP, we must find a solution s∗ ∈ S with maximum
(in the case of maximization) objective function value: f (s∗ ) ≥ f (s) ∀s ∈ S.
s∗ is the globally optimal solution of (S, f ), and the set S ∗ ⊆ S is the set of
globally optimal solutions.
There are many algorithms to deal with COPs, usually classified into two
categories: complete or approximate. Complete algorithms guarantee to find an
optimal solution in a time rate that grows as a polynomial function of the size of
the problem. On the other hand, approximate algorithms can obtain acceptable
solutions in a significantly reduced amount of time.
Among approximate algorithms, metaheuristics are tools that generate ap-
proximate solutions in reasonable times to complex COPs. They include one or
more heuristic methods using a higher-level strategy (hence ‘meta’). Heuristics
inside of a metaheuristic are considered a black-box as little (if any) prior knowl-
edge is known about it by the metaheuristic, so that it may be replaced with a
different heuristic [2].
On the other hand, the integration of information or knowledge from several
sources is a relevant problem at present, as many areas need to merge different
pieces of data to integrate a common and structured set of information, such
as expert opinion fusion in risk analysis, sensor fusion or logic [4]. A number of
approaches to tackle these problems have been developed. Among them, Belief
Merging is a technique based on model theory and propositional logic that has
been successfully applied on complex real-life problems such as IBM’s QRadar
exploit detection system [13] and cancer diagnosis [9].
There are hard COPs where a solely numerical approach can be used to solve
it, such as in optimization and search. However, some COPs may include the
integration of (possibly inconsistent) data, so we propose a high-level model to
manage the problem from two points of view: Metaheuristic optimization and
Belief merging.
The following core questions determine if our framework is suitable for a
given problem:
1. Is it a Combinatorial Optimization Problem?
If yes, a metaheuristic may fit the problem.
2. Does the problem need to combine different and conflicting information pieces?
If so, belief merging should be used in order to produce a single consistent
knowledge base which retains as much information as possible.
3. Can the problem be split into two sub-problems (numerical and symbolical)?
If it is the case, this means that two different approaches can be used to
solve the problem.
If the third question is affirmative, then we can employ our framework to
solve it.
52
The following include a background on metaheuristics and belief merging
formalisms (section 2), the description of our proposal (section 3), a motivating
example (section 4) and conclusions (section 5) .
2 Background
2.1 Metaheuristics
Recently, metaheuristics have gained popularity over mathematical program-
ming methods due to their flexibility and robust behavior to solve real-life COPs
[14]. Metaheuristics simulate or emulate processes or behavior inspired by na-
ture. A common classification include Evolutionary Algorithms (e.g., Genetic
Algorithms, Differential Evolution), Swarm Intelligence (e.g., Ant Colony Op-
timization, Bacterial Foraging Algorithm) and Global Search Algorithms (e.g.,
Simulated Annealing, Tabu Search) [12].
The family of Global Search Algorithms is among the most efficient meta-
heuristics. They manage a local search procedure (neighborhood exploring) in
order to combine exploration and exploitation conveniently. Local search algo-
rithms start from some initial solution and iteratively try to replace the current
solution with a better one in a defined neighborhood [1].
Definition 2. A neighborhood is a function N : S → 2S that assigns to every
s ∈ S a set of neighbors N (s) ⊆ S. N (s) is called the neighborhood of S.
Global Search Algorithms use the neighborhood structure along with a spe-
cific strategy to escape from local minima.
Definition 3. A local minimum solution w.r.t. a neighborhood N is a solution
s0 such that ∀s ∈ N (s0 ) : f (s0 ) < f (s).
The term Tabu Search was coined in the same paper that introduced the term
metaheuristic [8]. A distinguishing feature of this algorithm is the use of adaptive
memory and specific associated problem-solving strategies. Tabu Search provides
the origin of the memory-based and strategy-intensive focus in the metaheuristic
literature, as opposed to methods that are memory-less. It is also responsible for
emphasizing the use of structured designs to exploit historical patterns of search,
as opposed to processes that rely almost exclusively on randomization [5].
The particular characteristic of the Tabu Search is a short-term memory
called tabu list containing moves recently applied. This list is to disallow moves
that reverse the effect of recent moves, adding them a sort of forbidden status.
However, from time to time a tabu move can reach a better solution space. Thus,
aspiration criteria are implemented in order to revoke the tabu status of a move.
Many criteria have been proposed, but the most widely used aspiration criterion
consists in allowing a tabu move if it results in a solution with a better objective
value than the current best-known solution. It has been demonstrated in numer-
ous research and real-life projects that Tabu Search finds good approximations
to the optimal solution for large combinatorial problems [7].
53
In general, there are plenty of proposals on metaheuristic hybridization. The
most popular approach is a hybrid of Evolutionary Algorithm and a Local Search
Algorithm [12]. However, the combination of a metaheuristic with a symbolic
approach such as Belief Merging has not been proposed in the literature.
2.2 Belief merging
Belief merging is a knowledge fusion approach for the integration of knowledge
coming from different sources. This logic-based technique is defined as the oper-
ation of combining information from a set of (possibly conflicting) belief bases
obtained from different sources to produce a single and consistent belief base
[15].
In this particular case, each belief base corresponds to the knowledge or
preferences of any entity of the problem, e.g., stakeholder, sensor, robot, or
service. By merging these belief bases, we obtain a belief base that represents
the consensus of preferences, or the knowledge of the group.
Belief merging is based on propositional logic, with the significant number
of proposals based on model theory. So, in this theory we consider a language
L(P ) of propositional logic using a finite ordered set of symbols (atoms or vari-
ables) P = {p1 , p2 , ..., pn }. A propositional formula (or a set of formulas) ψ
is conformed by a subset of variables from P with the set of logic connectives
ρ = {¬, ∧, ∨, →, ↔} in the classical way. P(ψ) denotes the set of atoms appearing
in ψ. |P| denotes the cardinality of set P. A literal l is an atom or the negation
of an atom. A term D is a finite conjunction of literals i.e., D = l1 ∧ . . . ∧ lk ,
with li = pj or li = ¬pj . The Disjunctive Normal Form (DNFψ ) of every for-
mula ψ ∈ L(P ) is a disjunction of terms DNFψ =D1 ∨ . . . ∨ Dm , such that
ψ ≡ DNFψ . An interpretation is a function w from P to {1, 0} (the classical
truth values 1 representing true and 0 representing false). The set of all possible
interpretations is denoted as W, with elements denoted by vectors of the form
(w(p1 ), . . . , w(pn )). A model of ψ is an interpretation such that w(Q) = 1 once
w is extended in the usual way over the connectives, and the set of models of a
formula ψ will be denoted by mod (ψ). ψ is consistent iff there exists a model of
ψ.
Apart the classical definitions in model theory, we need the following defini-
tions [11]:
Definition 4. A belief base K is a finite set of formulas of L(P ) representing the
beliefs of a stakeholder. We also identify K with the conjunction of its elements,
for example if K = {ψ1 , ψ2 , . . . , ψn } then K = ψ1 ∧ ψ2 ∧ . . . ∧ ψn .
Definition 5. A belief profile E = {K1 , . . . , Km } denotes the beliefs of the group
of stakeholders involved in the fusion process. E is a multiset of m belief bases,
as two or more stakeholders may have identical beliefs.
Several belief merging operators have been defined and characterized in a
logical way. It is common to represent the result of a merging operator as a set
54
of interpretations instead of belief
bases, such that, if ∆ is an operator and E
a belief profile, then mod ∆(E) represents the set of interpretations resulting
from the fusion. However, it is possible to obtain the formula representing the
belief base from the set of interpretations, except equivalences.
Belief merging operators are categorized into two main families: model-based
operators and formula-based operators. Two subclasses of merging operators are
considered [10]:
1. Majority merging operators its goal is to satisfy the beliefs of all stake-
holders as a whole. If a subset of beliefs in the belief profile is repeated
enough times, then it is the one that prevails.
2. Arbitration merging operators focus on satisfying as much as possi-
ble each stakeholder. This subclass of operators selects the median possible
choices.
∆ps (PS-Merge) is a flexible operator classified as a majority-merging op-
erator but can be refined tending to be an arbitration operator. Unlike cited
operators, it can operate over inconsistent belief bases. It is not based on the
classic measure distance between models (Hamming distance), but in the notion
of Partial Satisfiability [16]. This notion allows savings at computing the models
of belief bases.
Notably, ∆ps refines the results of other Belief Merging operators in consen-
sus decision-making scenarios [17]. Thus, we propose ∆ps to conform the Belief
Merging module of our framework.
3 Hybrid framework
We divide our proposal into two modules, each one performing a task of the
problem solution. Each module uses different inputs. As significant data pre-
processing may be required, it may be an essential part of the framework. This
step may involve the creation of helper data structures, data cleaning or the
translation of data formats.
In the first phase, we suggest using a metaheuristic algorithm to solve the
optimization part of the problem, we propose specifically the Tabu Search.
The optimization part of the problem should be the first stage of the process,
as it produces a preliminary solution which in turn is used by the following
module.
For the second phase, a belief merging operator, specifically ∆ps , is used to
make the data fusion. At this stage, the input of this stage is the preliminary
solution built up in the previous stage. We need to translate this numerical
solution into a knowledge base format. This module also needs the other belief
bases to merge.
Figure 1 outlines our proposal.
55
Hybrid model
<> <> <>
Preprocessor Metahueristic engine Belief merger Solution
Metaheuristic parameters
Problem
data
Knowledge bases
Fig. 1. Hybrid framework high-level view.
3.1 Module 1: Numerical optimization engine
The optimization is performed by this module using Algorithm 1. This Tabu
Search implementation is based on the original tutorial by Fred Glover [6].
Main elements include:
Tabu list. It is a queue containing the disallowed moves. Its size, the tabu
tenure, determines the number of iterations that a move is classified as tabu.
Tabu tenure may be static or dynamic, depending on the problem type and
instance size.
Aspiration criterion. Conveniently called “strategic oblivion,” defines the re-
vocation way of a tabu status over a particular move, In this case, the con-
dition to override tabu status is to accept any move that reaches to a better
solution.
Initial solution. An initial solution should be provided in order to take advan-
tage of the deterministic nature of Tabu Search.
Stop criterion. There are many conditions to stop the optimization process.
Algorithm 1 ends according to one of the following criteria:
– Fixed number of iterations.
– Limit time, in seconds.
– Find a solution s∗ with a given objective function value.
– Number of iterations with no improved solution.
3.2 Module 2: Belief merger
This module has the following components:
Symbolic translator. A belief merging operator computes models of the lan-
guage, so this component converts the output of Module 1 into a knowledge
base format.
Definition 6. Let Sm1 be the partial solution generated by Module 1. Then,
translating each variable ϕ ∈ Sm1 into a logical variable ϕ̂ will give the belief
logic
base Sm1 = {ϕ̂1 , . . . , ϕ̂n }.
56
Algorithm 1: Tabu Search with simple aspiration criterion.
Input: Initial solution s
Output: Best solution s∗
1 k ← 1;
∗
2 s ← s;
3 while not stop criterion is met do
4 Identify N (s) ; // neighborhood
5 Identify T (s, k) ; // Tabu list
6 Identify A(s, k) ; // Aspiration set
7 V ← N (s, k) ⊆ N (s) − T (s, k) + A(s, k);
8 s ← best in V ;
9 if f (s) < f (s∗ ) then
10 s∗ ← s;
11 end
12 k++;
13 end
logic
Definition 7. The belief base Sm1 , along with the belief bases detected
logic
in the preprocessing stage will conform the knowledge profile E = Sm1 ∪
{K1 , . . . , Kn }.
Logical evaluator. This component evaluates the models of the belief profile.
It builds a truth table of size |P(E)| × 2|P(E)| .
To ease the evaluation of formulas, we transform each belief base into its RPN
(Reverse Polish Notation). We include the RPN using Dijkstra’s Shunting-
yard algorithm implemented for logical operators [3] using the operators
in ρ with the usual precedence. Where operators have equal precedence,
parenthesis along with their associativity indicates in which order they will
be applied.
Belief merging operator. Algorithm 2 presents an implementation of the ∆ps
operator using the notion of Partial Satisfiability, which is defined as follows:
Definition 8. Partial Satisfiability is defined over a belief base normalized
in DNF format: QK , any interpretation w ∈ W and |P| = n. The Partial
Satisfiability of QK for w, denoted by wps (QK ), is defined as:
1. If QK is a conjunction C1 ∧ . . . ∧ Cs where each Ci is a literal, then:
( s )
X w(Ci ) n − |P (Vs Ci )|
i=1
wps (QK ) = max , (1)
i=1
s 2n
where:
n is the number of atoms of the considered language.
s is the number of literals appearing in the conjunction.
57
Algorithm 2: ∆ps (PS-Merge)
Input : V : Number of variables of E
B: Number of bases of E
D: Vector of number of disjuncts of each base in E
L: Matrix of occurrences of literals in each disjunct of each base of E
Output: Solution Set: The set of models of ∆ps (E)
1 Solution ← ∅
2 M ax-Sum ← 0
3 W ← Matrix whose rows are all the possible interpretations for V variables.
4 for s = 1 . . . B do
IDs ← s−1
P
5 k=1 Dk+1
6 end
V
7 for i = 1 . . . 2 do
8 Sum ← 0
9 for s = 1 . . . B do
10 ps-disjunct ← ∅
11 for d = IDs . . . IDs + Ds do
12 satisf ied ← 0
13 conjuncts ← 0
14 vars-not-appearing ← 0
15 for j = 1 . . . V do
16 if Wi,j = 1 then
17 satisf ied ← satisf ied + Ld,2j−1
18 end
19 if Wi,j = 0 then
20 satisf ied ← satisf ied + Ld,2j
21 end
22 conjuncts ← conjuncts + Ld,2j−1 + Ld,2j
23 if Ld,2j = 0 and Ld,2j−1 = 0 then
24 vars-not-appearing ← vars-not-appearing+1
25 end
26 end
satisf ied vars−not−appearing
27 ps-disjunct ← ps-disjunct ∪{max( conjuncts , 2V
)}
28 end
29 P S ← max(ps − disjunct)
30 Sum ← Sum + P S
31 end
32 if Sum > M axSum then
33 Solution ← {i}
34 M axSum ← Sum
35 end
36 else if Sum = M axSum then
37 Solution ← Solution ∪ {i}
38 end
39 end
40 Solution Set = {ith row of W |i ∈ Solution}
58
2. If QK is a disjunction D1 ∨ . . . ∨ Dr where each Di is a conjunction of
literals, then:
wps (QK ) = max wps (D1 ), . . . , wps (Dr ) (2)
Definition 9. The following pre-order is defined:
m
X m
X
0 0
w ≤E
ps w iff wps (QKi ) ≤ wps (QKi ) (3)
i=1 i=1
Definition 10. ∆ps operator is defined by its models as:
mod ∆ps (E) = max W, ≤E
ps (4)
As mentioned, ∆ps belongs to the majority operators subclass. However, by
computing the minimum value from the Partial Satisfiability of the belief
bases, it tends to be an arbitration operator. With this refinement, we can
choose the resulting interpretations impartially, satisfying each stakeholder
as much as possible.
4 Motivating example
Many real-world COPs may include both numerical and symbolic issues. To illus-
trate our proposal, we now introduce the Course Timetabling Problem (CTP).
In general, CTP consists in the allocation of courses, groups of students and
professors to time slots, subject to constraints of two types:
– Hard constraints: Are conditions that must be fully satisfied in order to
obtain a feasible solution.
– Soft constraints: Are optional requirements that are desirable but not manda-
tory.
Hard constraints are considered requirements, while soft constraints are con-
sidered as preferences. As far as we know, there is no record of a proposal for
solution or hybrid method using Belief Merging.
There are various formulations of CTP [18], the following is a variation of
[19] considering the professors’ teaching preferences.
Example 1. A school needs a timetable of q courses K1 , . . . , Kq , and each i course
Ki consists of ki lectures. There are t teachers P1 , . . . , Pt , which all can teach any
lecture. There are r groups S1 , . . . , Sr of courses that have students in common.
This means that the courses in Sl must be scheduled all at different times. The
number of periods is p, and lk is the maximum number of lectures that can be
scheduled at period k (i.e., the number of rooms available at period k). It is
recommended to consider each teacher expertise, i.e., previous courses taught,
represented as a set Hj of courses {Ki1 , . . . , Kinj }. There is also an optional
list of courses preferences from each teacher, represented as a set Fj of courses
{Ki1 , ..., Kimj }.
59
4.1 Module 1
The following formulation serves as an input for the first module of our frame-
work, as it models the CTP as a search problem and therefore suitable for the
Tabu Search implemented:
find Sm1 = {xijk | i = 1, . . . , q; j = 1, . . . , t; k = 1, . . . , p} (5)
subject to the following constraints:
p
X
xijk = ki (i = 1, . . . , q; j = 1, . . . , t) (6)
k=1
Xq
xijk ≤ lk (j = 1, . . . , t; k = 1, . . . , p) (7)
i=1
X
xijk ≤ 1 (l = 1, . . . , r; j = 1, . . . , t; k = 1, . . . , p) (8)
i∈Sl
xijk ∈ {0, 1} with (i = 1, . . . , q; j = 1, . . . , t; k = 1, . . . , p) (9)
Sm1 , in this case, is a timetable where xijk = 1 if a lecture of course Ki given
by teacher Pj is scheduled at period k, and xijk = 0 otherwise. Constraint 6 im-
pose that each course is composed of the correct number of lectures. Constraint 7
enforce that at each time there aren’t more lectures than rooms. Constraint 8
prevent conflicting lectures to be scheduled during the same period. These are
the hard constraints of the problem.
Soft constraints are about the courses preferences. A pure optimization strat-
egy usually would address this soft constraint adding a weight variable dij rep-
resenting the desirability of courses by each teacher in a maximization objective
function. However, we will leave this task to the second module of our framework.
4.2 Module 2
A belief merging operator computes models of logic formulas, so this component
creates the belief bases by translating the crucial sets to make the fusion. Each
belief base represents a propositional formula where each element ϕ in the orig-
inal set is represented by the logical variable ϕ̂ This component translates the
following sets:
1. The timetable Sm1 generated by Module 1 into a knowledge bases format:
logic ^
Sm1 ,j = x̂ijk (∀xijk ∈ Sm1 | xijk = 1; i = 1, . . . , q; k = 1, . . . , p) (10)
2. The set of preferential courses into a knowledge base format:
logic ^
Fj = K̂ (∀K ∈ Fj ) (11)
60
3. A similar translation for the historical courses:
logic ^
Hj = K̂ (∀K ∈ Hj ) (12)
In order to fusion the desires and expertise into the current timetable, we
obtain the t independent belief profiles to merge, one for each teacher:
logic logic logic
Ej = Sm1 ,j , Fj , Hj (13)
∆ps evaluates the models of the belief profiles. The output will be the best
configuration of courses for each teacher, according to the soft constraints.
5 Conclusion
We present the preliminary ideas on the development of a hybrid framework for
Constraint Optimization Problems, taking advantages of two different techniques
such as Tabu Search and a Belief Merging operator.
We believe that this two-fold approach may produce more fair solutions in
problems where a consensus is needed, because that is precisely what Belief
Merging promotes, to satisfy most of the contradictory opinions.
Further work considers a full implementation of our proposal and tests over
real-world scenarios.
References
1. Blum, C., Roli, A.: Metaheuristics in combinatorial optimization: Overview
and conceptual comparison. ACM Comput. Surv. 35(3), 268–308 (2003).
https://doi.org/10.1145/937503.937505
2. Brownlee, J.: Clever algorithms : nature-inspired programming recipes. Lulu, On-
line (2011)
3. Chávez-Bosquez, O., Pozos-Parra, P., McAreavey, K.: On the development of a
logic calculator: a novel tool to perform logical operations. In: Osorio Galindo,
M., Marcial Romero, J., Zepeda Cortés, C., Olmos Pineda, I. (eds.) Tenth Latin
American Workshop on Logic/Languages, Algorithms and New Methods of Rea-
soning. pp. 9–16. CEUR Workshop Proceedings, Puebla, México (2016), http:
//ceur-ws.org/Vol-1659/paper2.pdf
4. Dubois, D., Liu, W., Ma, J., Prade, H.: The basic principles of un-
certain information fusion. an organised review of merging rules in dif-
ferent representation frameworks. Information Fusion 32, 12–39 (2016).
https://doi.org/http://dx.doi.org/10.1016/j.inffus.2016.02.006
5. Gendreau, M.: Handbook of Metaheuristics, chap. An Introduction to Tabu Search,
pp. 37–54. Springer US, Boston, MA (2003). https://doi.org/10.1007/0-306-48056-
52
6. Glover, F.: Tabu Search: A Tutorial. Interfaces 20(4), 21 (1990).
https://doi.org/10.1287/inte.20.4.74
7. Glover, F., Laguna, M.: Tabu Search. Kluwer Academic Publishers, USA, 1st edn.
(1997)
61
8. Glover, F.: Future paths for integer programming and links to artificial
intelligence. Computers & Operations Research 13(5), 533 – 549 (1986).
https://doi.org/https://doi.org/10.1016/0305-0548(86)90048-1, applications of In-
teger Programming
9. Kareem, S., Pozos-Parra, P., Wilson, N.: An application of belief merg-
ing for the diagnosis of oral cancer. Applied Soft Computing (2017).
https://doi.org/http://dx.doi.org/10.1016/j.asoc.2017.01.055
10. Konieczny, S., Pino Pérez, R.: Logic based merging. Journal of Philosophical Logic
40(2), 239–270 (2011)
11. Konieczny, S., Prez, R.P.: Merging information under constraints: A logi-
cal framework. Journal of Logic and Computation 12(5), 773–808 (2002).
https://doi.org/10.1093/logcom/12.5.773
12. Luke, S.: Essentials of Metaheuristics. Lulu, second edn. (2013), available for free
at http://cs.gmu.edu/∼sean/book/metaheuristics/
13. McAreavey, K., Liu, W., Miller, P.: Computational approaches to find-
ing and measuring inconsistency in arbitrary knowledge bases. Inter-
national Journal of Approximate Reasoning 55(8), 1659–1693 (2014).
https://doi.org/10.1016/j.ijar.2014.06.003
14. Michalewicz, Z., Fogel, D.: How to Solve It: Modern Heuristics. Springer-Verlag
Berlin Heidelberg, 2 edn. (2004). https://doi.org/10.1007/978-3-662-07807-5
15. Pigozzi, G.: Belief merging and judgment aggregation. In: Zalta, E.N. (ed.) The
Stanford Encyclopedia of Philosophy. Metaphysics Research Lab, Stanford Univer-
sity, winter 2016 edn. (2016)
16. Pozos Parra, P., Borja Macı́as, V.: Partial satisfiability-based merging. In: Gel-
bukh, A., Kuri Morales, Á. (eds.) MICAI 2007: Advances in Artificial Intelli-
gence. 6th Mexican International Conference on Artificial Intelligence. Lecture
Notes in Computer Science, vol. 4827, pp. 225–235. Springer Berlin Heidelberg
(2007). https://doi.org/10.1007/978-3-540-76631-5 22
17. Pozos-Parra, P., Chávez-Bosquez, O., McAreavey, K.: Merginator: A belief merging
tool for consensus support. Journal of Intelligent & Fuzzy Systems 34(5), 3199–
3210 (2018)
18. Schaerf, A.: A survey of automated timetabling. Artificial Intelligence Review
13(2), 87–127 (1999). https://doi.org/10.1023/A:1006576209967
19. de Werra, D.: An introduction to timetabling. European Journal of Operational
Research 19(2), 151–162 (1985). https://doi.org/10.1016/0377-2217(85)90167-5
62