• 정보보호학회논문지
• /
• 제33권6호
• /
• pp.1111-1123
• /
• 2023

최근 생성 모델이 발전함에 따라 생성 모델을 위협하는 연구도 활발히 진행되고 있다. 본 논문은 Text-to-Image 모델에 대한 멤버십 추론 공격을 위한 새로운 제안 방법을 소개한다. 기존의 Text-to-Image 모델에 대한 멤버십 추론 공격은 쿼리 이미지의 caption으로 단일 이미지를 생성하여 멤버십을 추론하였다. 반면, 본 논문은 Textual Inversion을 통해 쿼리 이미지에 personalization된 임베딩을 사용하고, Adversarial Prompt 생성 방법으로 여러 장의 이미지를 효과적으로 생성하는 멤버십 추론 공격을 제안한다. 또한, Text-to-Image 모델 중 주목받고 있는 Stable Diffusion 모델에 대한 멤버십 추론 공격을 최초로 진행하였으며, 최대 1.00의 Accuracy를 달성한다.

• 응용통계연구
• /
• 제34권3호
• /
• pp.389-399
• /
• 2021

최근 자연어 처리 문제에서의 단어 임베딩은 아주 큰 주목을 받고 있는 연구 주제이며 스킵그램은 성공적인 단어 임베딩 기법 중 하나이다. 주변단어들 정보를 이용해서 단어들의 의미를 학습하여 단어 임베딩 벡터를 할당하며 텍스트 자료를 효과적으로 분석할 수 있게 한다. 그러나 벡터 공간 모델의 한계로 인해 기본적인 단어 임베딩 방법들은 모든 단어가 하나의 의미를 가지고 있다는 것을 가정한다. 다의어, 즉 하나 이상의 의미를 가진 단어가 실생활에서 존재 하기 때문에 Neelakantan 등 (2014)은 군집분석 기법을 이용하여 다의어의 여러 의미들에 해당하는 의미 임베딩 벡터를 찾기 위해 MSSG (multi-sense skip-gram)를 제안했다. 본 논문에서는 MSSG의 통계적 성능을 개선시킬 수 있는 변형된 MSSG 방법을 제안한다. 먼저, 가중치를 활용한 가중문맥 벡터를 제안한다. 나아가, 군집의 수, 즉 다의어의 의미 수를 자료에서 자동적으로 추정해주는 x-means 방법을 활용한 알고리즘을 제안한다. 본 논문에서 수행한 실증자료를 기반한 모의실험에서 제안한 방법은 기존 방법에 비해 우수한 성능을 보여주었다.

• International Journal of Computer Science & Network Security
• /
• 제21권8호
• /
• pp.238-246
• /
• 2021

The rapid rise of the Internet and social media has resulted in a large number of text-based reviews being placed on sites such as social media. In the age of social media, utilizing machine learning technologies to analyze the emotional context of comments aids in the understanding of QoS for any product or service. The classification and analysis of user reviews aids in the improvement of QoS. (Quality of Services). Machine Learning algorithms have evolved into a powerful tool for analyzing user sentiment. Unlike traditional categorization models, which are based on a set of rules. In sentiment categorization, Bidirectional Long Short-Term Memory (BiLSTM) has shown significant results, and Convolution Neural Network (CNN) has shown promising results. Using convolutions and pooling layers, CNN can successfully extract local information. BiLSTM uses dual LSTM orientations to increase the amount of background knowledge available to deep learning models. The suggested hybrid model combines the benefits of these two deep learning-based algorithms. The data source for analysis and classification was user reviews of Indian Railway Services on Twitter. The suggested hybrid model uses the Keras Embedding technique as an input source. The suggested model takes in data and generates lower-dimensional characteristics that result in a categorization result. The suggested hybrid model's performance was compared using Keras and Word2Vec, and the proposed model showed a significant improvement in response with an accuracy of 95.19 percent.

• 대한산업공학회지
• /
• 제38권1호
• /
• pp.67-73
• /
• 2012

Identification of meaningful patterns and trends in large volumes of text data is an important task in various research areas. In the present study we crawled the keywords from the abstracts in IIE Transactions, one of the representative journals in the field of Industrial Engineering from 1969 to 2011. We applied low-dimensional embedding method, clustering analysis, association rule, and social network analysis to find meaningful associative patterns of key words frequently appeared in the paper.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권1호
• /
• pp.382-403
• /
• 2020

Steganography is the art of concealing the existence of a secret data in a non-secret digital carrier called cover media. While the image of steganography methods is extensively researched, studies on other cover files remain limited. Videos are promising research items for steganography primitives. This study presents an improved approach to video steganography. The improvement is achieved by allowing senders and receivers exchanging secret data without embedding the hidden data in the cover file as in traditional steganography methods. The method is based mainly on searching for exact matches between the secret text and the video frames RGB channel pixel values. Accordingly, a random key-dependent data is generated, and Elliptic Curve Public Key Cryptography is used. The proposed method has an unlimited embedding capacity. The results show that the improved method is secure against traditional steganography attacks since the cover file has no embedded data. Compared to other existing Steganography video systems, the proposed system shows that the method proposed is unlimited in its embedding capacity, system invisibility, and robustness. The system achieves high precision for data recovery in the receiver. The performance of the proposed method is found to be acceptable across different sizes of video files.

• International Journal of Computer Science & Network Security
• /
• 제22권7호
• /
• pp.352-358
• /
• 2022

Text mining is an important process used for analyzing the data collected from different sources like videos, audio, social media, and so on. The tools like Natural Language Processing (NLP) are mostly used in real-time applications. In the earlier research, text mining approaches were implemented using long-short memory (LSTM) networks. In this paper, text mining is performed using average-stochastic gradient descent weight-dropped (AWD)-LSTM techniques to obtain better accuracy and performance. The proposed model is effectively demonstrated by considering the internet movie database (IMDB) reviews. To implement the proposed model Python language was used due to easy adaptability and flexibility while dealing with massive data sets/databases. From the results, it is seen that the proposed LSTM plus weight dropped plus embedding model demonstrated an accuracy of 88.36% as compared to the previous models of AWD LSTM as 85.64. This result proved to be far better when compared with the results obtained by just LSTM model (with 85.16%) accuracy. Finally, the loss function proved to decrease from 0.341 to 0.299 using the proposed model

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권10호
• /
• pp.3211-3229
• /
• 2022

To solve the problems existing in the process of Weibo disaster rumor recognition, such as lack of corpus, poor text standardization, difficult to learn semantic information, and simple semantic features of disaster rumor text, this paper takes Sina Weibo as the data source, constructs a dataset for Weibo disaster rumor recognition, and proposes a deep learning model BERT_AT_Stacked LSTM for Weibo disaster rumor recognition. First, add adversarial disturbance to the embedding vector of each word to generate adversarial samples to enhance the features of rumor text, and carry out adversarial training to solve the problem that the text features of disaster rumors are relatively single. Second, the BERT part obtains the word-level semantic information of each Weibo text and generates a hidden vector containing sentence-level feature information. Finally, the hidden complex semantic information of poorly-regulated Weibo texts is learned using a Stacked Long Short-Term Memory (Stacked LSTM) structure. The experimental results show that, compared with other comparative models, the model in this paper has more advantages in recognizing disaster rumors on Weibo, with an F1_Socre of 97.48%, and has been tested on an open general domain dataset, with an F1_Score of 94.59%, indicating that the model has better generalization.

• Journal of information and communication convergence engineering
• /
• 제16권3호
• /
• pp.179-188
• /
• 2018

This paper proposes a quick response (QR) barcode watermarking algorithm in order to improve readability of the QR barcode in a JPEG image. The proposed algorithm embeds a watermark in a wavelet based-low frequency band for watermark robustness, and visually extracts a QR barcode after embedding the QR barcode image with confidential information into the original image via imperceptible watermarking technology. The visually extracted QR barcode watermark uses an approach to authenticate the ownership more easily through a common hardware and software-based mobile barcode reader app. Therefore, the QR barcode watermark may improve the ability to easily detect watermarks efficiently as well as imperceptibility and robustness, which is the main watermark requirement by embedding the QR barcode with watermark text information in a digital image and when compare to conventional watermarks.

• 한국정보기술학회 영문논문지
• /
• 제10권2호
• /
• pp.139-151
• /
• 2020

Entity name recognition is a part of information extraction that extracts entity names from documents and classifies the types of extracted entity names. Entity name recognition technologies are widely used in natural language processing, such as information retrieval, machine translation, and query response systems. Various deep learning-based models exist to improve entity name recognition performance, but studies that compared and analyzed these models on Korean data are insufficient. In this paper, we compare and analyze the performance of CRF, LSTM-CRF, BiLSTM-CRF, and BERT, which are actively used to identify entity names using Korean data. Also, we compare and evaluate whether embedding models, which are variously used in recent natural language processing tasks, can affect the entity name recognition model's performance improvement. As a result of experiments on patent data and Korean corpus, it was confirmed that the BiLSTM-CRF using FastText method showed the highest performance.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2021년도 추계학술발표대회
• /
• pp.494-497
• /
• 2021

텍스트 분류(text classification)는 원시 텍스트 데이터로부터 정보를 추출할 수 있는 기술에 기반하여 많은 양의 텍스트 데이터를 관심 영역으로 분류하는 것으로 최근에 각광을 받고 있다. 본 논문에서는 워드 임베딩(word embedding) 기법을 이용하여 특정 분야의 연구 논문을 분류하고 추천하는 기법을 제안한다. 워드 임베딩으로 CBOW(Continuous Bag-of-Word)와 Sg(Skip-gram)를 연구 논문의 분류에 적용하고 기존 방식인 TF-IDF(Term Frequency-Inverse Document Frequency)와 성능을 비교 분석한다. 성능 평가 결과는 워드 임베딩에 기반한 연구 논문 분류 기법이 TF-IDF에 기반한 연구 논문 분류 기법보다 좋은 성능을 가진다는 것을 나타낸다.