scholarly journals Avaliando atributos para a classificação de estrutura retórica em resumos científicos

Linguamática ◽  
2019 ◽  
Vol 11 (1) ◽  
pp. 41-53
Author(s):  
Alessandra Harumi Iriguti ◽  
Valéria Delisandra Feltrim
Keyword(s):  
Support Vector Machines ◽  
Decision Trees ◽  
Random Fields ◽  
Conditional Random Fields ◽  
Naive Bayes ◽  
Nearest Neighbors ◽  
Support Vector ◽  
Word Embeddings ◽  
K Nearest Neighbors ◽  
Vector Machines

A classificação de estrutura retórica é uma tarefa de PLN na qual se busca identificar os componentes retóricos de um discurso e seus relacionamentos. No caso deste trabalho, buscou-se identificar automaticamente categorias em nível de sentenças que compõem a estrutura retórica de resumos científicos. Especificamente, o objetivo foi avaliar o impacto de diferentes conjuntos de atributos na implementação de classificadores retóricos para resumos científicos escritos em português. Para isso, foram utilizados atributos superficiais (extraídos como valores TF-IDF e selecionados com o teste chi-quadrado), atributos morfossintáticos (implementados pelo classificador AZPort) e atributos extraídos a partir de modelos de word embeddings (Word2Vec, Wang2Vec e GloVe, todos previamente treinados). Tais conjuntos de atributos, bem como as suas combinações, foram usados para o treinamento de classificadores usando os seguintes algoritmos de aprendizado supervisionado: Support Vector Machines, Naive Bayes, K-Nearest Neighbors, Decision Trees e Conditional Random Fields (CRF). Os classificadores foram avaliados por meio de validação cruzada sobre três corpora compostos por resumos de teses e dissertações. O melhor resultado, 94% de F1, foi obtido pelo classificador CRF com as seguintes combinações de atributos: (i) Wang2Vec--Skip-gram de dimensões 100 com os atributos provenientes do AZPort; (ii) Wang2Vec--Skip-gram e GloVe de dimensão 300 com os atributos do AZPort; (iii) TF-IDF, AZPort e embeddings extraídos com os modelos Wang2Vec--Skip-gram de dimensões 100 e 300 e GloVe de dimensão 300. A partir dos resultados obtidos, conclui-se que os atributos provenientes do classificador AZPort foram fundamentais para o bom desempenho do classificador CRF, enquanto que a combinação com word embeddings se mostrou válida para a melhoria dos resultados.

scholarly journals Pengembangan Algoritma C4.5 Berbasis Particle Swarm Optimization untuk Penentuan Kelayakan Asuransi

2021 ◽  
Vol 2 (3) ◽  
pp. 427-437
Author(s):  
Jatmiko Indriyanto ◽  
Miftakhul Huda ◽  
Ida Afriliana
Keyword(s):  
Particle Swarm Optimization ◽  
Support Vector Machines ◽  
Decision Trees ◽  
Roc Curve ◽  
Particle Swarm ◽  
Nearest Neighbors ◽  
Support Vector ◽  
K Nearest Neighbors ◽  
Swarm Optimization ◽  
Vector Machines

Tujuan penelitian ini adalah untuk Pengembangan Algoritma C4.5 Berbasis Particle Swarm Optimization Untuk Penentuan Kelayakan Asuransi. metode logistic regresion, decision trees, k-nearest neighbors, naïve bayes dan support vector machines.  Model tersebut akan menentukan atau memprediksi status konsumen dimasa mendatang. Observasi yang mirip juga pernah dilakukan, tetapi dengan cara berbeda. Pada penilitian ini, akan digunakan algoritma klasifikasi C4.5 berbasis Particle Swarm Optimization (PSO), hasil ketepatan yang diinginkan lebih bagus dibandingkan hanya memakai algoritma C4.5 untuk mengatasi permasalahan pada kasus pemilihan produk asuransi. dapat disimpulkan bahwa nilai akurasi yang didapatkan pada model algoritma C4.5 berbasis PSO adalah 98.93% lebih bagus jika dibandingkan dengan model algoritma C4.5 yaitu 97.84%. Dari hasil tersebut didapatkan perbedaan antara kedua model yaitu senilai 0.4%. Selagi untuk penelaahan menggunakan ROC curve bagi kedua model ialah,untuk model algoritma C4.5 nilai AUC adalah 0.970 dengan urutan diagnosa  Excellent Classification, dan untuk model algoritma C4.5 berbasis PSO nilai AUC adalah  0.968 dengan urutan diagnosa Excellent Classification.

scholarly journals An Enhanced Corpus for Arabic Newspapers Comments

2020 ◽  
Vol 17 (5) ◽  
pp. 789-798
Author(s):  
Hichem Rahab ◽  
Abdelhafid Zitouni ◽  
Mahieddine Djoudi
Keyword(s):  
Support Vector Machines ◽  
Web Sites ◽  
Naive Bayes ◽  
Nearest Neighbors ◽  
Naïve Bayes ◽  
Support Vector ◽  
K Nearest Neighbors ◽  
Vector Machines

In this paper, we propose our enhanced approach to create a dedicated corpus for Algerian Arabic newspapers comments. The developed approach has to enhance an existing approach by the enrichment of the available corpus and the inclusion of the annotation step by following the Model Annotate Train Test Evaluate Revise (MATTER) approach. A corpus is created by collecting comments from web sites of three well know Algerian newspapers. Three classifiers, support vector machines, naïve Bayes, and k-nearest neighbors, were used for classification of comments into positive and negative classes. To identify the influence of the stemming in the obtained results, the classification was tested with and without stemming. Obtained results show that stemming does not enhance considerably the classification due to the nature of Algerian comments tied to Algerian Arabic Dialect. The promising results constitute a motivation for us to improve our approach especially in dealing with non Arabic sentences, especially Dialectal and French ones

Persian Handwritten Number Recognition Using Adapted Framing Feature and Support Vector Machines

2016 ◽  
Vol 15 (01) ◽  
pp. 1650004 ◽  
Author(s):  
Hedieh Sajedi ◽  
Mehran Bahador
Keyword(s):  
Support Vector Machines ◽  
Recognition Rate ◽  
Nearest Neighbors ◽  
Polynomial Kernel ◽  
Support Vector ◽  
K Nearest Neighbors ◽  
New Approach ◽  
Number Recognition ◽  
Vector Machines

In this paper, a new approach for segmentation and recognition of Persian handwritten numbers is presented. This method utilizes the framing feature technique in combination with outer profile feature that we named this the adapted framing feature. In our proposed approach, segmentation of the numbers into digits has been carried out automatically. In the classification stage of the proposed method, Support Vector Machines (SVM) and k-Nearest Neighbors (k-NN) are used. Experimentations are conducted on the IFHCDB database consisting 17,740 numeral images and HODA database consisting 102,352 numeral images. In isolated digit level on IFHCDB, the recognition rate of 99.27%, is achieved by using SVM with polynomial kernel. Furthermore, in isolated digit level on HODA, the recognition rate of 99.07% is achieved by using SVM with polynomial kernel. The experiments illustrate that applying our proposed method resulted higher accuracy compared to previous researches.

A Semantic Scattering model for the automatic interpretation of English genitives

2009 ◽  
Vol 15 (2) ◽  
pp. 215-239 ◽  
Author(s):  
ADRIANA BADULESCU ◽  
DAN MOLDOVAN
Keyword(s):  
Support Vector Machines ◽  
Decision Trees ◽  
Naive Bayes ◽  
Word Sense Disambiguation ◽  
Naïve Bayes ◽  
Semantic Relations ◽  
Support Vector ◽  
Word Sense ◽  
Vector Machines ◽  
Bayes Algorithm

AbstractAn important problem in knowledge discovery from text is the automatic extraction of semantic relations. This paper addresses the automatic classification of thesemantic relationsexpressed by English genitives. A learning model is introduced based on the statistical analysis of the distribution of genitives' semantic relations in a corpus. The semantic and contextual features of the genitive's noun phrase constituents play a key role in the identification of the semantic relation. The algorithm was trained and tested on a corpus of approximately 20,000 sentences and achieved an f-measure of 79.80 per cent for of-genitives, far better than the 40.60 per cent obtained using a Decision Trees algorithm, the 50.55 per cent obtained using a Naive Bayes algorithm, or the 72.13 per cent obtained using a Support Vector Machines algorithm on the same corpus using the same features. The results were similar for s-genitives: 78.45 per cent using Semantic Scattering, 47.00 per cent using Decision Trees, 43.70 per cent using Naive Bayes, and 70.32 per cent using a Support Vector Machines algorithm. The results demonstrate the importance of word sense disambiguation and semantic generalization/specialization for this task. They also demonstrate that different patterns (in our case the two types of genitive constructions) encode different semantic information and should be treated differently in the sense that different models should be built for different patterns.

Detection of Loss Zones while Drilling Using Different Machine Learning Techniques

2021 ◽  
pp. 1-29
Author(s):  
Ahmed Alsaihati ◽  
Mahmoud Abughaban ◽  
Salaheldin Elkatatny ◽  
Abdulazeez Abdulraheem
Keyword(s):  
Machine Learning ◽  
Support Vector Machines ◽  
Random Forests ◽  
Nearest Neighbors ◽  
Machine Learning Techniques ◽  
Support Vector ◽  
K Nearest Neighbors ◽  
Learning Techniques ◽  
Vector Machines ◽  
Testing Set

Abstract Fluid loss into formations is a common operational issue that is frequently encountered when drilling across naturally or induced fractured formations. This could pose significant operational risks, such as well-control, stuck pipe, and wellbore instability, which, in turn, lead to an increase of well time and cost. This research aims to use and evaluate different machine learning techniques, namely: support vector machines, random forests, and K-nearest neighbors in detecting loss circulation occurrences while drilling using solely drilling surface parameters. Actual field data of seven wells, which had suffered partial or severe loss circulation, were used to build predictive models, while Well-8 was used to compare the performance of the developed models. Different performance metrics were used to evaluate the performance of the developed models. Recall, precision, and F1-score measures were used to evaluate the ability of the developed model to detect loss circulation occurrences. The results showed the K-nearest neighbors classifier achieved a high F1-score of 0.912 in detecting loss circulation occurrence in the testing set, while the random forests was the second-best classifier with almost the same F1-score of 0.910. The support vector machines achieved an F1-score of 0.83 in predicting the loss circulation occurrence in the testing set. The K-nearest neighbors outperformed other models in detecting the loss circulation occurrences in Well-8 with an F1-score of 0.80. The main contribution of this research as compared to previous studies is that it identifies losses events based on real-time measurements of the active pit volume.

scholarly journals A Clinical Decision Support Tool to Detect Invasive Ductal Carcinoma in Histopathological Images Using Support Vector Machines, Naïve-Bayes, and K-Nearest Neighbor Classifiers

2020 ◽  
Author(s):  
Kyra Mikaela M. Lopez ◽  
Ma. Sheila A. Magboo
Keyword(s):  
Support Vector Machines ◽  
Invasive Ductal Carcinoma ◽  
Naive Bayes ◽  
Ductal Carcinoma ◽  
Naïve Bayes ◽  
Machine Learning Techniques ◽  
Support Vector ◽  
K Nearest Neighbors ◽  
Support Tool ◽  
Vector Machines

This study aims to describe a model that will apply image processing and traditional machine learning techniques specifically Support Vector Machines, Naïve-Bayes, and k-Nearest Neighbors to identify whether or not a given breast histopathological image has Invasive Ductal Carcinoma (IDC). The dataset consisted of 54,811 breast cancer image patches of size 50px x 50px, consisting of 39,148 IDC negative and 15,663 IDC positive. Feature extraction was accomplished using Oriented FAST and Rotated BRIEF (ORB) descriptors. Feature scaling was performed using Min-Max Normalization while K-Means Clustering on the ORB descriptors was used to generate the visual codebook. Automatic hyperparameter tuning using Grid Search Cross Validation was implemented although it can also accept user supplied hyperparameter values for SVM, Naïve Bayes, and K-NN models should the user want to do experimentation. Aside from computing for accuracy, the AUPRC and MCC metrics were used to address the dataset imbalance. The results showed that SVM has the best overall performance, obtaining accuracy = 0.7490, AUPRC = 0.5536, and MCC = 0.2924.

scholarly journals Maschinelles Lernen mit Aussagen zur Modellkompetenz

2020 ◽  
Vol 26 (1) ◽  
pp. 157-172
Author(s):  
Dirk Krüger ◽  
Moritz Krell
Keyword(s):  
Logistic Regression ◽  
Support Vector Machines ◽  
Decision Trees ◽  
Naive Bayes ◽  
Naïve Bayes ◽  
Support Vector ◽  
Maschinelles Lernen ◽  
Vector Machines

ZusammenfassungVerfahren des maschinellen Lernens können dazu beitragen, Aussagen in Aufgaben im offenen Format in großen Stichproben zu analysieren. Am Beispiel von Aussagen von Biologielehrkräften, Biologie-Lehramtsstudierenden und Fachdidaktiker*innen zu den fünf Teilkompetenzen von Modellkompetenz (NTraining = 456; NKlassifikation = 260) wird die Qualität maschinellen Lernens mit vier Algorithmen (naïve Bayes, logistic regression, support vector machines und decision trees) untersucht. Evidenz für die Validität der Interpretation der Kodierungen einzelner Algorithmen liegt mit zufriedenstellender bis guter Übereinstimmung zwischen menschlicher und computerbasierter Kodierung beim Training (345–607 Aussagen je nach Teilkompetenz) vor, bei der Klassifikation (157–260 Aussagen je nach Teilkompetenz) reduziert sich dies auf eine moderate Übereinstimmung. Positive Korrelationen zwischen dem kodierten Niveau und dem externen Kriterium Antwortlänge weisen darauf hin, dass die Kodierung mit naïve Bayes keine gültigen Ergebnisse liefert. Bedeutsame Attribute, die die Algorithmen bei der Klassifikation nutzen, entsprechen relevanten Begriffen der Niveaufestlegungen im zugrunde liegenden Kodierleitfaden. Abschließend wird diskutiert, inwieweit maschinelles Lernen mit den eingesetzten Algorithmen bei Aussagen zur Modellkompetenz die Qualität einer menschlichen Kodierung erreicht und damit für Zweitkodierungen oder in Vermittlungssituationen genutzt werden könnte.

Sign in / Sign up

Export Citation Format

Share Document