Professional Documents
Culture Documents
Textbook Hybrid Artificial Intelligent Systems Francisco Javier de Cos Juez Ebook All Chapter PDF
Textbook Hybrid Artificial Intelligent Systems Francisco Javier de Cos Juez Ebook All Chapter PDF
https://textbookfull.com/product/hybrid-intelligent-systems-
ajith-abraham/
https://textbookfull.com/product/hybrid-intelligent-systems-19th-
international-conference-on-hybrid-intelligent-systems-
his-2019-held-in-bhopal-india-december-10-12-2019-ajith-abraham/
https://textbookfull.com/product/hybrid-artificial-intelligent-
systems-14th-international-conference-hais-2019-leon-spain-
september-4-6-2019-proceedings-hilde-perez-garcia/
https://textbookfull.com/product/nature-inspired-design-of-
hybrid-intelligent-systems-1st-edition-patricia-melin/
https://textbookfull.com/product/artificial-intelligence-and-
algorithms-in-intelligent-systems-radek-silhavy/
https://textbookfull.com/product/artificial-intelligence-systems-
based-on-hybrid-neural-networks-theory-and-applications-michael-
zgurovsky/
https://textbookfull.com/product/advanced-electrocatalysts-for-
low-temperature-fuel-cells-francisco-javier-rodriguez-varela/
Francisco Javier de Cos Juez
José Ramón Villar · Enrique A. de la Cal
Álvaro Herrero · Héctor Quintián
José António Sáez
Emilio Corchado (Eds.)
LNAI 10870
Hybrid Artificial
Intelligent Systems
13th International Conference, HAIS 2018
Oviedo, Spain, June 20–22, 2018
Proceedings
123
Lecture Notes in Artificial Intelligence 10870
Hybrid Artificial
Intelligent Systems
13th International Conference, HAIS 2018
Oviedo, Spain, June 20–22, 2018
Proceedings
123
Editors
Francisco Javier de Cos Juez Héctor Quintián
Department of Mine Operating University of A Coruña
and Prospection A Coruña
University of Oviedo Spain
Oviedo
Spain José António Sáez
University of Salamanca
José Ramón Villar Salamanca
Department of Computer Science Spain
University of Oviedo
Oviedo Emilio Corchado
Spain University of Salamanca
Salamanca
Enrique A. de la Cal Spain
Department of Computer Science
University of Oviedo
Oviedo
Spain
Álvaro Herrero
Department of Civil Engineering
University of Burgos
Burgos
Spain
This Springer imprint is published by the registered company Springer International Publishing AG
part of Springer Nature
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
General Chair
Emilio Corchado University of Salamanca, Spain
Local Chairs
Francisco Javier University of Oviedo, Spain
de Cos Juez
Jose Ramón Villar University of Oviedo, Spain
Enrique A. de la Cal University of Oviedo, Spain
Honorary Committee
Javier Fernández Fernández President of the Government of Asturias
Santiago García Granda Chancellor of the University of Oviedo
Francisco Wenceslao Mayor of the City of Oviedo
López Martínez
Program Committee
Enrique de la Cal University of Oviedo, Spain
(PC Co-chair)
Francisco Javier de Cos University of Oviedo, Spain
Juez (PC Co-chair)
Héctor Quintián University of A Coruña, Spain
(PC Co-chair)
José Ramón Villar University of Oviedo, Spain
(PC Co-chair)
Emilio Corchado (PC Chair) University of Salamanca, Spain
Abdel-Badeeh Salem Ain Shams University, Egypt
Alberto Cano Virginia Commonwealth University, USA
Alfredo Cuzzocrea ICAR-CNR and University of Calabria, Italy
Alicia Troncoso Pablo de Olavide University, Spain
Álvaro Herrero University of Burgos, Spain
Amelia Zafra Gómez University of Córdoba, Spain
Ana M. Bernardos Polytechnic University of Madrid, Spain
Ana Madureira Instituto Superior de Engenharia do Porto, Portugal
Anca Andreica Babes-Bolyai University, Romania
Andreea Vescan Babes-Bolyai University, Romania
Andrés Enrique University of Oviedo, Spain
Andrés Pinón University of A Coruña, Spain
Ángel Arroyo University of Burgos, Spain
Antonio de Jesús Díez University of Oviedo, Spain
Antonio D. Masegosa University of Deusto/IKERBASQUE, Spain
Antonio Dourado University of Coimbra, Portugal
Antonio Morales-Esteban University of Seville, Spain
Arkadiusz Kowalski Wrocław University of Technology, Poland
Barna Laszlo Iantovics Petru Maior University of Tg. Mures, Romania
Beatriz Remeseiro University of Oviedo, Spain
Bogdan Trawinski Wroclaw University of Science and Technology,
Poland
Organization IX
Organizing Committee
Enrique de la Cal University of Oviedo, Spain
José R. Villar University of Oviedo, Spain
Francisco Javier University of Oviedo, Spain
de Cos Juez
Noelia Rico University of Oviedo, Spain
Mirko Fáñez University of Oviedo, Spain
Carmen Peñalver University of Oviedo, Spain
Juan Manuel Marín University of Oviedo, Spain
Marta Blanco University of Oviedo, Spain
Héctor Quintian University of A Coruña, Spain
José Antonio Sáez University of Salamanca, Spain
Emilio Corchado University of Salamanca, Spain
Contents
SmartFD: A Real Big Data Application for Electrical Fraud Detection. . . . . . 120
D. Gutiérrez-Avilés, J. A. Fábregas, J. Tejedor, F. Martínez-Álvarez,
A. Troncoso, A. Arcos, and J. C. Riquelme
Tuning CNN Input Layout for IDS with Genetic Algorithms . . . . . . . . . . . . 197
Roberto Blanco, Juan J. Cilla, Pedro Malagón, Ignacio Penas,
and José M. Moya
Learning Algorithms
Dealing with Missing Data and Uncertainty in the Context of Data Mining . . . 289
Aliya Aleryani, Wenjia Wang, and Beatriz De La Iglesia
Using Nonlinear Quantile Regression for the Estimation of Software Cost . . . 422
J. De Andrés, M. Landajo, and P. Lorca
Memetic Modified Cuckoo Search Algorithm with ASSRS for the SSCF
Problem in Self-Similar Fractal Image Reconstruction . . . . . . . . . . . . . . . . . 658
Akemi Gálvez, Andrés Iglesias, Iztok Fister, Iztok Fister Jr.,
Eneko Osaba, and Javier Del Ser
1 Introduction
knowledge, an architecture for a knowledge base for the financial domain in the Linked
Open Data (LOD) cloud is presented in this paper. As stated by the DBPedia project,
knowledge bases are playing an increasingly important role in enhancing the intelli-
gence of Web and in supporting information integration.
Furthermore, recommendation systems have over the years proved to be effective in
overcoming the challenges related to the incredible growth of the information on the
Web. In this context, deep learning techniques, which can exploit massive amounts of
data, have recently gained much attention from recommender systems researchers after
companies such as Yahoo, YouTube and Google bet on these techniques in an attempt
to improve their recommendation quality [2].
From the assumption that these facts are true also in the case of the huge amounts of
financial knowledge in the LOD cloud, we proposed a deep learning-based hybrid
recommendation system as the means to enable end user access to the knowledge base.
One of the advantages of our knowledge base compared with others in the literature
is the integration of a recommendation system, which eliminates the need for the users
to known the underlying ontologies and use both an ontology language and a graph
query language. At the same time, this integration makes the proposed recommendation
system different from others in the literature. In fact, the results of the state-of-the-art
analysis carried out in this research suggest that recommendation systems have not yet
been sufficiently studied as a means to support knowledge bases in the Linked Data
cloud.
An Information Systems-oriented evaluation based on traditional Information
Retrieval metrics is used for the validation of the presented architecture.
The remainder of this paper is structured as follows: the fundamentals of the topics
of this research are described in Sect. 2. The architecture that is the salient contribution
of this paper is outlined in Sect. 3. Section 4 presents the evaluation method used for
validation purposes. Finally, conclusions are discussed in Sect. 5.
2 Background
sources, (2) Linked Open Data (LOD) datasets from DBPedia (http://wiki.dbpedia.org/)
and crunchbase (https://www.crunchbase.com/) and (3) news media.
Furthermore, a method for publishing and linking financial data, which is called
“Financial Linked Data (FLD)”, was presented in [5]. This method uses the XBRL
standard to link facts to corresponding business reports and then publish these facts
using Linked Data principles. FLD also allows interlinking resulting financial data with
relevant external LOD datasets such as the Corporate and Individual Ownership
dataset.
The building of ontologies is key to the materialization of the vision of the
Semantic Web that is enabled by the Linked Data principles. Despite many proposals
for the automatic construction and maintenance of ontologies have arisen in the last few
years [6–8], these tasks are considered a pending issue in ontology engineering.
In this context, it is possible to distinguish three major categories of automatic
ontology building tasks: ontology learning and ontology evolution, which are beyond
the scope of this work, and ontology population. A particular case of ontology pop-
ulation that specifically concerns this work is ontology population from semi-structured
documents, such as XML, RSS and HTML documents.
Ontology population from semi-structured HTML documents has traditionally
received much attention from ontology engineering researches because most of the
content on the Web is provided by means of HTML-based Web pages in which tables
can be seen as the natural means to structure information about entities [9].
Various approaches for populating ontologies from semi-structured HTML docu-
ments that focus on HTML tables have been proposed to date. Nonetheless, a general
methodology for this task can involve the following phases: (1) crawl the Web in
search of HTML documents relevant to the domain of interest; (2) use Web wrappers to
process the tables within the HTML documents and extract information about occur-
rences of domain entities by regarding each table entry as an occurrence of an unknown
entity given by a set of pairs of a possible property name and the corresponding value;
(3) interpret each property name in a set as a predicate of an RDF triple, and each value
as an object of this predicate in the same triple; (3.1) match the domain entity repre-
sented by the table as a whole with the most proper concept in the ontology; interpret
the resulting concept as the subject common to all the outlined RDF triples [10].
As proved by companies like Yahoo, YouTube and Google, deep learning tech-
niques significantly improved traditional model-based CF techniques [13–15], which
are probably the most widely-used techniques in recommender systems at present [16].
Moreover, deep learning techniques can be applied to problems involving massive
amounts of data, which is the case of problems related to access and consumption of
knowledge from knowledge bases in the Linked Data cloud.
In this context, an approach to learning top-N recommendations from knowledge
graphs, which is called “entity2Rec”, was presented in [17]. In that work, graphs are feed
with linked open data, and a global user-item relatedness model, which is computed from
vector representations of the graph nodes, is proposed by relying on the “Adarank” and
“LambdaMART” learning to rank algorithms. Node vector representations are computed
using “node2vec”, a deep feature learning framework for networks, which is considered
an improvement over the “DeepWalk” algorithm [18].
Furthermore, in [19] a unified framework for collaborative filtering and knowledge
base embedding, which is called “Collaborative Knowledge Base Embedding (KBE)” is
proposed to simultaneously learn items’ semantic representations from the knowledge
base and capture the implicit relations between users and items. Semantic representations
are automatically extracted by using deep learning embedding techniques.
Nonetheless, unlike these proposals, in this work the deep learning-based recom-
mendation techniques are not the end but the means to support a knowledge base.
3 Proposed Architecture
At a high level of abstraction, the architecture for the knowledge base for the financial
domain, which is the salient contribution of this paper, is composed of five major
components: (1) Domain Ontology, (2) Ontology Population Subsystem, (3) Knowledge
Repository, (4) Semantic Annotation Subsystem and (5) Recommendation Subsystem.
(see Fig. 1). The roles and associations between all these components are explained
below.
Where R+(u) are the items (iʹ) liked by the active user u in the past, s is the adjusted
cosine similarity metric, xp(u), xp(i) and xp(iʹ) are the property-specific vector
representations.
10 L. O. Colombo-Mendoza et al.
4 Evaluation
Table 1. Results of Recall, Precision and F1 score calculation for the proposed approach.
Relevant Non-relevant Recom. Recall Prec. F1 score
User System-hits User System-errors
14 10 6 2 12 0.833 0.714 0.769
12 9 8 2 11 0.818 0.750 0.783
13 11 7 3 14 0.786 0.846 0.815
14 10 6 4 14 0.714 0.714 0.714
13 11 7 4 15 0.733 0.846 0.786
14 10 6 4 14 0.714 0.714 0.714
13 10 7 3 13 0.769 0.769 0.769
12 9 8 2 12 0.818 0.750 0.783
12 10 8 3 13 0.769 0.833 0.800
11 9 9 3 12 0.750 0.818 0.783
Average 0.771 0.776 0.772
Table 2 depicts the results of the comparative analysis carried out during the second
stage of the evaluation. These results are given in terms of average Recall, Precision
and F1 score values.
According to the results from Table 1, the average Recall value of the proposed
recommendation approach is 0.771 (77.1%), whereas its average Precision value is
0.776 (77.6%). In particular, in the best-case scenario in terms of Precision, 11 out of
Another random document with
no related content on Scribd:
femme, abandonne Olympie. — Roger reçoit
l’hippogriffe des mains de Logistilla qui lui apprend à le
conduire. Il descend avec lui en Angleterre, où il voit le
rassemblement des troupes destinées à porter secours
à Charles. En passant en Irlande, il aperçoit dans l’île
d’Ébude Angélique enchaînée à un rocher pour être
dévorée par l’orque. Il abat le monstre, prend la jeune
fille en croupe, et descend avec elle sur le rivage de la
Basse-Bretagne.
Chant XI. — Angélique échappe à Roger au moyen de
l’anneau enchanté, et se réfugie dans la demeure d’un
pasteur. Roger, allant à sa recherche, voit un géant
enlever une dame qui lui paraît être Bradamante. —
Olympie, abandonnée par Birène et prise par des
corsaires, est exposée dans l’île d’Ébude à la voracité
du monstre marin. Roland la délivre. Survient Obert,
roi d’Irlande, qui devient amoureux d’Olympie et la
prend pour femme, après avoir enlevé à Birène ses
États et la vie. 219
Chant XII. — Roland, toujours à la recherche
d’Angélique, voit une femme qui lui ressemble dans les
bras d’Atlante, lequel, changé en chevalier, paraît
l’emporter. En le poursuivant, Roland parvient à un
palais enchanté, où arrive également Roger qui court
après celui qu’il prend pour le ravisseur de
Bradamante. Angélique y arrive, elle aussi, et y trouve
Roland, Sacripant, Ferragus, Gradasse et d’autres
guerriers. Une querelle s’élève à son sujet entre
quelques-uns d’entre eux, ce qui procure à Ferragus
l’occasion de s’emparer du casque de Roland.
Angélique se dirige vers le Levant, et trouve dans un
bois un jeune homme mortellement blessé. — Roland
va vers Paris et détruit deux troupes de Maures. Plus
loin il découvre un repaire de malandrins qui retiennent
Isabelle prisonnière. 240
Chant XIII. — Isabelle raconte à Roland ses malheurs. 264
Surviennent les malandrins habitants de la caverne.
Roland les tue tous, puis il part emmenant Isabelle. —
Bradamante apprend de Mélisse que Roger est tombé
au pouvoir du vieux magicien. Elle va pour le délivrer
et reste prise dans son propre enchantement. —
Digression élogieuse de Mélisse sur les femmes
appartenant à la maison d’Este.
Chant XIV. — L’armée des païens s’étant rassemblée, on
constate l’absence des deux troupes détruites par
Roland. Mandricard, courant sur les traces du paladin,
rencontre Doralice, fille du roi de Grenade, qui s’en va
épouser Rodomont, roi de Sarze. Il tue le cortège,
emmène Doralice avec lui et en fait sa femme. Les
Maures donnent l’assaut à Paris. 284
Notes 319
NOTES DU TRANSCRIPTEUR
Les notes numérotées de [1] à [15] sont les notes de
bas de page de la préface. On a numéroté de [16] à [67]
les notes en fin de volume, qui ne présentent dans l’original
ni numérotation ni renvoi depuis le corps du texte.
*** END OF THE PROJECT GUTENBERG EBOOK ROLAND
FURIEUX, TOME 1 ***
1.D. The copyright laws of the place where you are located also
govern what you can do with this work. Copyright laws in most
countries are in a constant state of change. If you are outside
the United States, check the laws of your country in addition to
the terms of this agreement before downloading, copying,
displaying, performing, distributing or creating derivative works
based on this work or any other Project Gutenberg™ work. The
Foundation makes no representations concerning the copyright
status of any work in any country other than the United States.
1.E.6. You may convert to and distribute this work in any binary,
compressed, marked up, nonproprietary or proprietary form,
including any word processing or hypertext form. However, if
you provide access to or distribute copies of a Project
Gutenberg™ work in a format other than “Plain Vanilla ASCII” or
other format used in the official version posted on the official
Project Gutenberg™ website (www.gutenberg.org), you must, at
no additional cost, fee or expense to the user, provide a copy, a
means of exporting a copy, or a means of obtaining a copy upon
request, of the work in its original “Plain Vanilla ASCII” or other
form. Any alternate format must include the full Project
Gutenberg™ License as specified in paragraph 1.E.1.
• You pay a royalty fee of 20% of the gross profits you derive from
the use of Project Gutenberg™ works calculated using the
method you already use to calculate your applicable taxes. The
fee is owed to the owner of the Project Gutenberg™ trademark,
but he has agreed to donate royalties under this paragraph to
the Project Gutenberg Literary Archive Foundation. Royalty
payments must be paid within 60 days following each date on
which you prepare (or are legally required to prepare) your
periodic tax returns. Royalty payments should be clearly marked
as such and sent to the Project Gutenberg Literary Archive
Foundation at the address specified in Section 4, “Information
about donations to the Project Gutenberg Literary Archive
Foundation.”
• You comply with all other terms of this agreement for free
distribution of Project Gutenberg™ works.
1.F.
Most people start at our website which has the main PG search
facility: www.gutenberg.org.