Journal of Intelligent & Fuzzy Systems - Volume 39, issue 2

Authors: Meng, Fei | Wei, Jianliang

Article Type: Research Article

Abstract: With the promotion of opinion leader’s impact on online purchase intention, the problem of how to measure the characteristics of opinion leader, the characteristics of opinion leader’s recommendation information and the influence of consumers’ characteristics on purchase intention is becoming more and more urgent. Based on numbers of popular scales, this paper designs the questionnaire items for the variables of professional knowledge, product involvement, visual cues, interactivity, functional value and trust involved in the opinion leader influence model, and forms the initial scale. On this basis, with the help of small-scale interviews, small sample pre-test and large sample test, trust …and purchase intention fail to pass the validity test. Through correlation coefficient analysis, some questions with lower coefficient value are eliminated, and then the final scale with good reliability and validity is obtained. Show more

Keywords: Opinion leader, purchase intention, scale design, questionnaire

DOI: 10.3233/JIFS-179964

Citation: Journal of Intelligent & Fuzzy Systems, vol. 39, no. 2, pp. 1937-1949, 2020

Authors: Wang, Qinge | Chen, Huihua

Article Type: Research Article

Abstract: In order to overcome the problems of long execution time and low parallelism of existing parallel random forest algorithms, an optimization method for parallel random forest algorithm based on distance weights is proposed. The concept of distance weights is introduced to optimize the algorithm. Firstly, the training sample data are extracted from the original data set by random selection. Based on the extracted results, a single decision tree is constructed. The single decision tree is grouped together according to different grouping methods to form a random forest. The distance weights of the training sample data set are calculated, and then …the weighted optimization of the random forest model is realized. The experimental results show that the execution time of the parallel random forest algorithm after optimization is 110 000 ms less than that before optimization, and the operation efficiency of the algorithm is greatly improved, which effectively solves the problems existing in the traditional random forest algorithm. Show more

Keywords: Distance weights, parallel algorithm, random forest algorithm, algorithm optimization

DOI: 10.3233/JIFS-179965

Citation: Journal of Intelligent & Fuzzy Systems, vol. 39, no. 2, pp. 1951-1963, 2020

Authors: Liu, Ying

Article Type: Research Article

Abstract: At present, the teaching of architectural art in China is still relatively traditional, and there are still some problems in the actual teaching. Based on this, this study combines the Naive Bayesian classification algorithm with the fuzzy model to construct a new architectural art teaching model. In teaching, the Naive Bayesian classification algorithm generates only a small number of features for each item in the training set, and it only uses the probability calculated in the mathematical operation to train and classify the item. Moreover, by combining the fuzzy model, the materials needed for architectural art teaching can be quickly …generated, and the teaching principles and implementation strategies of architectural art are summarized. In addition, this paper proposes an attribute weighted classification algorithm combining differential evolution algorithm with Naive Bayes. The algorithm assigns weights to each attribute based on the Naive Bayesian classification algorithm and uses differential evolution algorithm to optimize the weights. The research shows that the method proposed in this paper has certain effect on the optimization of architectural art teaching mode. Show more

Keywords: Bayesian classification algorithm, fuzzy model, architectural art, differential evolution algorithm

DOI: 10.3233/JIFS-179966

Citation: Journal of Intelligent & Fuzzy Systems, vol. 39, no. 2, pp. 1965-1976, 2020

Article Type: Other

Citation: Journal of Intelligent & Fuzzy Systems, vol. 39, no. 2, pp. 1977-1977, 2020

Authors: Navas-Loro, María | Rodríguez-Doncel, Víctor

Article Type: Research Article

Abstract: Temporal information is crucial in knowledge extraction. Being able to locate events in a timeline is necessary to understand the narrative behind every text. To this aim, several temporal taggers have been proposed in literature –nevertheless, not all languages received the same attention. Most taggers work only for English texts, and not many have been developed for other languages. Also the scarcity of annotated corpora in other languages notably hinders the task. In this paper we present a new rule-based tagger called Annotador (Añotador in Spanish) able to process texts both in Spanish and English. Furthermore, a new …corpus with more than 300 short texts containing common temporal expressions, called the HourGlass corpus, has been built in order to test it and to facilitate the development of new resources and tools. Professionals from different domains intervened in the gathering of the text, making it heterogeneous and easy to use thanks to the tags added to each entry. Finally, we analyzed main challenges in the time expression extraction task. Show more

Keywords: Time expression, temporal tagger, Spanish language, NLP

DOI: 10.3233/JIFS-179865

Citation: Journal of Intelligent & Fuzzy Systems, vol. 39, no. 2, pp. 1979-1991, 2020

Authors: Kolesnikova, Olga | Gelbukh, Alexander

Article Type: Research Article

Abstract: In this work, we report the results of our experiments on the task of distinguishing the semantics of verb-noun collocations in a Spanish corpus. This semantics was represented by four lexical functions of the Meaning-Text Theory. Each lexical function specifies a certain universal semantic concept found in any natural language. Knowledge of collocation and its semantic content is important for natural language processing, as collocation comprises the restrictions on how words can be used together. We experimented with word2vec embeddings and six supervised machine learning methods most commonly used in a wide range of natural language processing tasks. Our objective …was to study the ability of word2vec embeddings to represent the context of collocations in a way that could discriminate among lexical functions. A difference from previous work with word embeddings is that we trained word2vec on a lemmatized corpus after stopwords elimination, supposing that such vectors would capture a more accurate semantic characterization. The experiments were performed on a collection of 1,131 Excelsior newspaper issues. As the experimental results showed, word2vec representation of collocations outperformed the classical bag-of-words context representation implemented in a vector space model and fed into the same supervised learning methods. Show more

Keywords: Word embeddings, word2vec, supervised machine learning, lexical function, Meaning-Text Theory

DOI: 10.3233/JIFS-179866

Citation: Journal of Intelligent & Fuzzy Systems, vol. 39, no. 2, pp. 1993-2001, 2020

Authors: Millán-Hernández, Christian Eduardo | García-Hernández, René Arnulfo | Ledeneva, Yulia | Hernández-Castañeda, Ángel

Article Type: Research Article

Abstract: A drug name could be confused because it looks or sounds like another. Nevertheless, it is not possible to know a priori the causes of the confusion. Nowadays, sophisticated similarity measures have been proposed focused on improving the score of the detection. However, when a new drug name is proposed, the Federal Drug Administration (FDA) only can reject or accept the drug name based on this value. This paper not only improves the detection of confused drug names by integrating the strengths of different similarity measures but also the orthographic and phonetic knowledge of these measures are used to give …an a priori explanation of the causes of confusion. In this paper, a novel measure that integrates 24 individual measures is developed for this problem. With our proposal, each individual measure contributes to this problem. Finally, we present examples of how our proposal is used for explaining the causes of the confusion which could assist to the FDA to accept or reject a new drug name or to know the confusion causes of previously reported cases. Show more

Keywords: LASA error, knowledge-based similarity measure, confused drug names, orthographic measure, phonetic measure

DOI: 10.3233/JIFS-179867

Citation: Journal of Intelligent & Fuzzy Systems, vol. 39, no. 2, pp. 2003-2013, 2020

Authors: Ramos-Flores, Orlando | Pinto, David | Montes-y-Gómez, Manuel | Vázquez, Andrés

Article Type: Research Article

Abstract: This work presents an experimental study on the task of Named Entity Recognition (NER) for a narrow domain in Spanish language. This study considers two approaches commonly used in this kind of problem, namely, a Conditional Random Fields (CRF) model and Recurrent Neural Network (RNN). For the latter, we employed a bidirectional Long Short-Term Memory with ELMO’s pre-trained word embeddings for Spanish. The comparison between the probabilistic model and the deep learning model was carried out in two collections, the Spanish dataset from CoNLL-2002 considering four classes under the IOB tagging schema, and a Mexican Spanish news dataset with seventeen …classes under IOBES schema. The paper presents an analysis about the scalability, robustness, and common errors of both models. This analysis indicates in general that the BiLSTM-ELMo model is more suitable than the CRF model when there is “enough” training data, and also that it is more scalable, as its performance was not significantly affected in the incremental experiments (by adding one class at a time). On the other hand, results indicate that the CRF model is more adequate for scenarios having small training datasets and many classes. Show more

Keywords: Named entity recognition, CRF, Bi-LSTM, Spanish, news reports

DOI: 10.3233/JIFS-179868

Citation: Journal of Intelligent & Fuzzy Systems, vol. 39, no. 2, pp. 2015-2025, 2020

Authors: Millán-Hernández, Christian Eduardo | García-Hernández, René Arnulfo | Ledeneva, Yulia

Article Type: Research Article

Abstract: Since a drug name goes through different communication means and circumstances when it is prescribed, written, advertised, listened to, searched and administered; it tends to be confused with similar drug names that Look-Alike and Sound-Alike (LASA). LASA drug names have caused costs and damage to health. For this problem, the institutions of the United Kingdom, Canada, and the United States have implemented programs for several decades to report lists of confusing drug names pairs. Thanks to these kinds of list, it has been possible to propose new models to identify confusing drug names in English and are used to reject …new drug name proposals or to alert when a confusing drug name is being dispensed. However, countries such as Spain also have published a list with the Spanish LASA drug names, and it is not clear enough whether the models previously proposed for the drug names in English are useful for the list in Spanish or if it is necessary to adjust and update them for the Spanish language. This paper focuses on updating and improving the identification of LASA drug names in Spanish. First, we update the state-of-the-art by evaluating all the individual similarity measures proposed previously and all the models that combine these measures with the list in Spanish. Second, we updated the models with new individual measures and then adjusted them with the list in Spanish to improve the identification of LASA drug names in Spanish. After that, 25 individual similarity measures and 8 models to identify confused drug names in Spanish are compared to obtain the best result and conclusions. Show more

Keywords: Look-alike and sound-alike drug names, spanish LASA problem, similarity measures, combined similarity measures

DOI: 10.3233/JIFS-179869

Citation: Journal of Intelligent & Fuzzy Systems, vol. 39, no. 2, pp. 2027-2036, 2020

Authors: Acharya, Harshith R. | Bhat, Aditya D. | Avinash, K. | Srinath, Ramamoorthy

Article Type: Research Article

Abstract:  In this paper, we propose the LegoNet - a system to classify and summarize legal judgments using Sentence Embedding, Capsule Networks and Unsupervised Extractive Summarization. To train and test the system, we have created a mini-corpus of Indian legal judgments which have been annotated according to the classes: Facts, Arguments, Evidences and Judgments. The proposed framework uses Sentence Embedding and Capsule Networks to classify parts of legal judgments into the classes mentioned above. This is then used by the extractive summarizer to generate a concise and succinct summary of the document grouped according to the above mentioned classes. Such …a system could be used to help enable the Legal Community by speeding up the processes involving reading and summarizing legal documents which a Law professional would undertake in preparing for a case. The performance of the Machine Learning Model in this architecture can improve over time as more annotated training data is added to the corpus. Show more

Keywords: Law Domain, Capsule Network, Sentence Embedding, Unsupervised Extractive Summarization, Natural Language Processing, Text Classification

DOI: 10.3233/JIFS-179870

Citation: Journal of Intelligent & Fuzzy Systems, vol. 39, no. 2, pp. 2037-2046, 2020