|Unsupervised Topic Labeling of Text Based on Wikipedia Categorization|
Defining text topicality is often an expensive problem that
requires significant resources for text labeling. Though many
packages already exist that provide dictionaries of labeled text,
synonyms, and Part-of-Speach tagging, the problem is ongoing
as language develops and new meanings of words and phrases
emerge. This paper proposes a cheap in human labor solution to
topic labeling of any text in the majority of languages. The
methodology uses links to the naturally emerging corpus of
labeled text – the Wikipedia. Wikipedia categories are
processed to extract a weighted set of topic labels for the
analyzed text. The approach is evaluated by processing
categorized texts and comparing the similarity of the top ranks
of topic labels to the text category. The topic labels extracted
using this methodology can be used for comparing similarity of
texts, for the assessment of the completeness of topic coverage
in automated marking of essays, and for coding in qualitative
text analysis. The paper contributes to the field of NLP by
offering a cheap and organically developing method of topical
text labeling. The paper contributes to the work of qualitative
analysts by offering a methodology for the analysis of interview
transcripts and other unstructured text.
The Outlines of an Art Machine
Tirtha Prasad Mukhopadhyay, Victor Hugo Jimenez
In this paper we propose to examine the cognitive aspects of
artistic creation. Art objects are supposed to elicit emotional
responses in the viewer. Behavior related to the making of art
objects are analysed. Both visual art and artistic verbal
expressions are considered for analysis. Emotional appraisal is
claimed to be indispensable to artistic creativity, as opposed to
appraisal objectives in design cognition where structural
variation and the resulting innovations produced could well be
emotively neutral in their appearance. The authors propose a
heuristic and connective-functionalist thesis of machine art
following identification of responsive elements for art as they are
laid down in precepts of different philosophical traditions. The
insights deriving from ancient and contemporary traditions
demonstrate that innovative variation in art presupposes the
presence of a set of corresponding variations in visual patterns or
linguistic expressions that typify a range of expectations for
target objects. A database of categorically defined ‘genre’ of art
should exhibit visual or verbal preferences in interactions.
Binary operations may be domain specific depending on the kind
of art that is under scrutiny, but from a philosophical perspective,
emotional representation must be assumed to be indispensable
across generic requirements.
Use of Artificial Intelligence in Supply Chain Management Practices and 3PL Selection
Aicha Aguezzoul, Silvio Pires
This study is focused on discuss the application of Artificial Intelligence (AI) techniques in the general case of SCM practices, on the one hand, and in the particular case of the 3PL selection process, on the other hand.
Concerning the SCM, the main purpose is to identify how current knowledge in AI could contribute to and be used effectively in SCM, especially in the conduction of its more dynamic managerial practices. In the case of 3PL selection process, the objective is to identify the proposed AI techniques used, taking into account the business sector of the company, and the logistics services that the company plans to outsource.
Holistic Development of Undergraduate Students – Concept Cartoons to Authentic Discovery
Kausik S. Das, Larry Gonick, Monica Mitchell, Charles G. Baldwin, Moses Kairo
This paper describes a holistic pedagogical approach for classroom engagement. The project translates theory and fundamental classroom knowledge to authentic application with cutting edge research implemented by undergraduates at a Historically Black University. In our project, we developed and assessed cartoons custom designed for classroom instruction and evaluated student engagement while using the cartoons. We further report on student successes achieved through undergraduate research projects.
Short-Range Wind Speed Predictions in Subtropical Region Using Artificial Intelligence
Pedro Junior Zucatelli, Erick Giovani Sperandio Nascimento, Alejandro Mauricio Gutiérrez Arce, Davidson Martins Moreira
Short-range wind speed predictions for subtropical region is performed by applying Artificial Neural Network (ANN) technique to the hourly time series representative of the site. To train the ANN and validate the technique, data for one year are collected by one tower, with anemometers installed at heights of 101.8, 81.8, 25.7, and 10.0 m. Different ANN configurations to Multilayer Perceptron (MLP), Recurrent Neural Network (RNN), Gated Recurrent Unit (GRU), and Long Short-Term Memory (LSTM), a deep learning algorithm based method, are applied for each site and height. A quantitative analysis is conducted and the statistical results are evaluated to select the configuration that best predicts the real data. These methods have lower computational costs than other techniques, such as numerical modelling. The proposed method is an important scientific contribution for reliable large-scale wind power forecasting and integration into existing grid systems in Uruguay. The best results of the short-term wind speed forecasting was for MLP, which performed the forecasts using a hybrid method based on recursive inference, followed by LSTM, at all the anemometer heights tested, suggesting that this method is a powerful tool that can help the Administración Nacional de Usinas y Transmissiones Eléctricas manage the national energy supply.
An Empirical Analysis of the Influence of Seismic Data Modeling for Estimating Velocity Models with Fully Convolutional Networks
Luan Rios Campos, Peterson Nogueira, Davidson Moreira, Erick Giovani Sperandio Nascimento
Seismic modeling is the process of simulating wave propagations in a medium to represent underlying structures of a subsurface area of the earth. This modeling is based on a set of parameters that determine how the data is produced. Recent studies have demonstrated that deep learning methods can be trained with seismic data to estimate velocity models that give a representation of the subsurface where the seismic data was generated. Thus, an analysis is made on the impact that different sets of parameters have on the estimation of velocity models by a fully convolutional network (FCN). The experiments varied the number of sources among four options (1, 10, 25 or 50 shots) and used three different ranges of peak frequencies: 4, 8 and 16 Hz. The results demonstrated that, although the number of sources have more influence on the computational time needed to train the FCN than the peak frequency, both changes have significant impact on the quality of the estimation. The best estimations were obtained with the experiment of 25 sources with 4 Hz and increasing the peak frequency to 8 Hz improved even more the results, especially regarding the FCN’s loss function.
|Artificial Intelligence in Medicine: Preparing for the Confirmed Inevitable. Theoretical and Methodological Considerations|
Andrey V. Rezaev, Piotr K. Yablonskiy
The AI in Medicine project began with a simple yet complex and multilevel question. In late 2017, prompted by direct experience of researching human-machine interchanges, we asked whether the traditional principles of interaction between a physician and a patient in the time of technological and computer revolution had changed. That, in turn, led to other questions. Was the very concept of principles of doctor-patient interaction, as an interaction between ‘Subject’ and ‘Object’, still relevant in the 21st century? While such principles are not deterministic, in the past they were followed meticulously. Whether they still wield their original instructive power is an intriguing question. But it is not our immediate purpose. We do not intend to replace one set of principles, locked up to time and place, with another set equally constrained. We acknowledge that there would be no quick and easy answers. As an initial move we simply seek to elicit the right questions.
We hope our paper will offer a mechanism for constructive engagement, discussion and discovery. The broadest possible engagement is crucial to meeting the kaleidoscope of irregular issues in interactions between medical professionals and general public that characterizes our time of Internet dominance.
More importantly, the paper extends an invitation to think anew, across the traditional barriers of scholarly disciplines, policies and habits.
A Survey on Computational Intelligence Techniques in User Identity Management
Abhijit Kumar Nag
User identity is a critical part to secure the legitimate access of authentication to protect his personal sensitive information. In recent years, computational intelligence (CI) plays a significant role to innovate various authentication approaches to enhance this identity management. In this comprehensive survey paper, various computational identity verification techniques are discussed and how these techniques provide enhanced security in different levels are highlighted. Moreover, an empirical comparison of various authentication techniques is illustrated in this paper to reflect the overall current research directions in the areas of computational intelligence in the user identity management system.
Studying Artificial Intelligence and Artificial Sociality in Natural Sciences,
Engineering, and Social Sciences: Possibility and Reality
Andrey V. Rezaev, Anastasia A. Ivanova
The paper highlights issues of studying artificial intelligence
(AI). The path taken here is to engage the reader in a discussion
of interdisciplinarity/crossdisciplinarity of AI studies. It begins
with a basic assumption and key argument that antidisciplinarity
rather than inter- or multi-disciplinarity will bring
a new dynamic to scientific research dealing with “artificial
intelligence” and “artificial sociality”. Discussion of the social
scientists’ concerns and problems is reported in what follows.
On this base the authors develop their ideas which may help
theorists and empirical researchers to tackle questions of AI
development in a society. In a conclusion the paper makes
correlations of the research outcomes with a reality of higher
Maximum Power Point Tracking Method Based on Perturb and Observe Coupled with a Neural Network for Photovoltaic Systems Operating Under Fast Changing Environments
Yesid Briceno-Fajardo, Gustavo Cerda-Villafana, Sergio Ledesma-Orozco
The output power of Photovoltaic (PV) arrays presents a
nonlinear behavior. Its maximum power point varies with the
cell’s temperature and solar radiation. It is due to this situation
that Maximum Power Point Tracking (MPPT) methods have
been proposed and used in order to maximize the PV array
output power. This paper presents an artificial neural network
(ANN) combined with the classic Perturbation and Observation
(P&O) algorithm to accelerate the search of such Maximum
Power Point. Simulations generated using Matlab/Simulink
show the improvement compared to the P&O alone and the
hardware implementation, using a 16-bit microcontroller
corroborates these findings.
Road State Classification of Bangladesh with Convolutional Neural Network Approach
Sajid Ahmed, Taoseef Ishtiak, Arif Ur Rahaman Chowdhury Suhan, Mehreen Hossain Anila, Tanjila Farah
The Traffic congestion is one of the most intricate and challenging problems in all major cities and urban area of Bangladesh. Inadequate road infrastructure is one of the major causes involved with this agonizing issue. The only existing solution to this issue is manual reporting to authority. This study proposes an app-based road state classification, damage detection, and reporting system to assist both the drivers and authority to identify the damaged roads through a proposed web platform. This paper has made various contributions to address the road type classification of Bangladesh. The proposed research work includes the first of its kind road surface classification dataset, prepared in Bangladesh that could be used for applying machine learning techniques. The dataset has been classified in five classes based on the surface condition. The research team then studied some of the state-of-the-art Residual network based machine learning models and later proposed a customized architecture with a smaller number of layers compared to the state-of-the-art Inception-v3 and Inception-ResNet-V2 architectures for classification purpose. The study has explored three different state-of-the-art machine learning models i.e. Inception-v3, Inception-ResNet-v2, Xception for classification and analyzed their results.