• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.41 no.8
• /
• pp.958-964
• /
• 2016

In this paper, we propose an i-vector based language recognition system to identify the spoken language of the speaker, which uses multiple discriminative deep neural network (DNN) models analogous to the multi-class support vector machine (SVM) classification system. The proposed model was trained and tested using the i-vectors included in the NIST 2015 i-vector Machine Learning Challenge database, and shown to outperform the conventional language recognition methods such as cosine distance, SVM and softmax NN classifier in open-set experiments.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.12
• /
• pp.3868-3888
• /
• 2022

A widely used social networking service like Twitter has the ability to disseminate information to large groups of people even during a pandemic. At the same time, it is a convenient medium to share irrelevant and unverified information online and poses a potential threat to society. In this research, conventional machine learning algorithms are analyzed to classify the data as either non-rumor data or rumor data. Machine learning techniques have limited tuning capability and make decisions based on their learning. To tackle this problem the authors propose a deep learning-based Rumor Detection Neural Network model to predict the rumor tweet in real-world events. This model comprises three layers, AttCNN layer is used to extract local and position invariant features from the data, AttBi-LSTM layer to extract important semantic or contextual information and HPOOL to combine the down sampling patches of the input feature maps from the average and maximum pooling layers. A dataset from Kaggle and ground dataset #gaja are used to train the proposed Rumor Detection Neural Network to determine the veracity of the rumor. The experimental results of the RDNN Classifier demonstrate an accuracy of 93.24% and 95.41% in identifying rumor tweets in real-time events.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.19 no.1
• /
• pp.119-126
• /
• 2024

In building information model (BIM), it is difficult to train an artificial intelligence (AI) model due to the lack of sufficient data about individual projects in an architecture firm. In this paper, we present a methodology to correctly train an AI neural network model based on a large language model (LLM) to predict the steel structure product weight ratios in BIM. The proposed method, with the aid of the LLM, can overcome the inherent problem of limited data availability in BIM and handle a combination of natural language and numerical data. The experimental results showed that the proposed method demonstrated significantly higher accuracy than methods based on a smaller language model. The potential for effectively applying large language models in BIM is confirmed, leading to expectations of preventing building accidents and efficiently managing construction costs.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.11
• /
• pp.441-448
• /
• 2019

Previous researches on automatic spacing of Korean sentences has been researched to correct spacing errors by using n-gram based statistical techniques or morpheme analyzer to insert blanks in the word boundary. In this paper, we propose an end-to-end automatic word spacing by using deep neural network. Automatic word spacing problem could be defined as a tag classification problem in unit of syllable other than word. For contextual representation between syllables, Bi-LSTM encodes the dependency relationship between syllables into a fixed-length vector of continuous vector space using forward and backward LSTM cell. In order to conduct automatic word spacing of Korean sentences, after a fixed-length contextual vector by Bi-LSTM is classified into auto-spacing tag(B or I), the blank is inserted in the front of B tag. For tag classification method, we compose three types of classification neural networks. One is feedforward neural network, another is neural network language model and the other is linear-chain CRF. To compare our models, we measure the performance of automatic word spacing depending on the three of classification networks. linear-chain CRF of them used as classification neural network shows better performance than other models. We used KCC150 corpus as a training and testing data.

• Journal of KIISE
• /
• v.44 no.2
• /
• pp.179-185
• /
• 2017

Document summarization aims to generate a summary that is consistent and contains the highly related sentences in a document. In this study, we implemented for document summarization that extracts highly related sentences from a whole document by considering both similarities and entailment relations between sentences. Accordingly, we proposed a new algorithm, TextRank-NLI, which combines a Recurrent Neural Network based Natural Language Inference model and a Graph-based ranking algorithm used in single document extraction-based summarization task. In order to evaluate the performance of the new algorithm, we conducted experiments using the same datasets as used in TextRank algorithm. The results indicated that TextRank-NLI showed 2.3% improvement in performance, as compared to TextRank.

• ETRI Journal
• /
• v.39 no.4
• /
• pp.455-466
• /
• 2017

This study focuses on a method for sequential data augmentation in order to alleviate data sparseness problems. Specifically, we present corpus expansion techniques for enhancing the coverage of a language model. Recent recurrent neural network studies show that a seq2seq model can be applied for addressing language generation issues; it has the ability to generate new sentences from given input sentences. We present a method of corpus expansion using a sentence-chain based seq2seq model. For training the seq2seq model, sentence chains are used as triples. The first two sentences in a triple are used for the encoder of the seq2seq model, while the last sentence becomes a target sequence for the decoder. Using only internal resources, evaluation results show an improvement of approximately 7.6% relative perplexity over a baseline language model of Korean text. Additionally, from a comparison with a previous study, the sentence chain approach reduces the size of the training data by 38.4% while generating 1.4-times the number of n-grams with superior performance for English text.

• Journal of Convergence for Information Technology
• /
• v.9 no.8
• /
• pp.1-8
• /
• 2019

Dialog system is becoming a new dominant interaction way between human and computer. It allows people to be provided with various services through natural language. The dialog system has a common structure of a pipeline consisting of several modules (e.g., speech recognition, natural language understanding, and dialog management). In this paper, we tackle a task of domain classification for the natural language understanding module by employing machine learning models such as convolutional neural network and random forest. For our dataset of seven service domains, we showed that the random forest model achieved the best performance (F1 score 0.97). As a future work, we will keep finding a better approach for domain classification by investigating other machine learning models.

• KIISE Transactions on Computing Practices
• /
• v.22 no.3
• /
• pp.127-132
• /
• 2016

Several methods have been proposed to apply artificial neural networks to text compression in the past. However, the networks and targets are both limited to the small size due to hardware capability in the past. Modern GPUs have much better calculation capability than CPUs in an order of magnitude now, even though CPUs have become faster. It becomes possible now to train greater and complex neural networks in a shorter time. This paper proposed a method to transform the distribution of original data with a probabilistic neural predictor. Experiments were performed on a feedforward neural network and a recurrent neural network with gated-recurrent units. The recurrent neural network model outperformed feedforward network in compression rate and prediction accuracy.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.12
• /
• pp.4706-4724
• /
• 2020

With various structured data, such as the company size, loan balance, and savings accounts, the voice of customer (VOC), which is text data containing contact history and counseling details was analyzed in this study. To analyze unstructured data, the term frequency-inverse document frequency (TF-IDF) analysis, semantic network analysis, sentiment analysis, and a convolutional neural network (CNN) were implemented. A performance comparison of the models revealed that the predictive model using the CNN provided the best performance with regard to predictive power, followed by the model using the TF-IDF, and then the model using semantic network analysis. In particular, a character-level CNN and a word-level CNN were developed separately, and the character-level CNN exhibited better performance, according to an analysis for the Korean language. Moreover, a systematic selection model for optimal text mining techniques was proposed, suggesting which analytical technique is appropriate for analyzing text data depending on the context. This study also provides evidence that the results of previous studies, indicating that individual customers leave when their loyalty and switching cost are low, are also applicable to corporate customers and suggests that VOC data indicating customers' needs are very effective for predicting their behavior.

• International Journal of Computer Science & Network Security
• /
• v.23 no.9
• /
• pp.120-128
• /
• 2023

In this work, we have presented ATSA, a hierarchical attention deep learning model for Arabic sentiment analysis. ATSA was proposed by addressing several challenges and limitations that arise when applying the classical models to perform opinion mining in Arabic. Arabic-specific challenges including the morphological complexity and language sparsity were addressed by modeling semantic composition at the Arabic morphological analysis after performing tokenization. ATSA proposed to perform phrase-chunks sentiment embedding to provide a broader set of features that cover syntactic, semantic, and sentiment information. We used phrase structure parser to generate syntactic parse trees that are used as a reference for ATSA. This allowed modeling semantic and sentiment composition following the natural order in which words and phrase-chunks are combined in a sentence. The proposed model was evaluated on three Arabic corpora that correspond to different genres (newswire, online comments, and tweets) and different writing styles (MSA and dialectal Arabic). Experiments showed that each of the proposed contributions in ATSA was able to achieve significant improvement. The combination of all contributions, which makes up for the complete ATSA model, was able to improve the classification accuracy by 3% and 2% on Tweets and Hotel reviews datasets, respectively, compared to the existing models.