• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.10
• /
• pp.2643-2657
• /
• 2023

Scene graphs are structured representations that can clearly convey objects and the relationships between them, but are often heavily biased due to the highly skewed, long-tailed relational labeling in the dataset. Indeed, the visual world itself and its descriptions are biased. Therefore, Unbiased Scene Graph Generation (USGG) prefers to train models to eliminate long-tail effects as much as possible, rather than altering the dataset directly. To this end, we propose Geometric and Semantic Improvement (GSI) for USGG to mitigate this issue. First, to fully exploit the feature information in the images, geometric dimension and semantic dimension enhancement modules are designed. The geometric module is designed from the perspective that the position information between neighboring object pairs will affect each other, which can improve the recall rate of the overall relationship in the dataset. The semantic module further processes the embedded word vector, which can enhance the acquisition of semantic information. Then, to improve the recall rate of the tail data, the Class Balanced Seesaw Loss (CBSLoss) is designed for the tail data. The recall rate of the prediction is improved by penalizing the body or tail relations that are judged incorrectly in the dataset. The experimental findings demonstrate that the GSI method performs better than mainstream models in terms of the mean Recall@K (mR@K) metric in three tasks. The long-tailed imbalance in the Visual Genome 150 (VG150) dataset is addressed better using the GSI method than by most of the existing methods.

• ETRI Journal
• /
• v.43 no.1
• /
• pp.95-108
• /
• 2021

Application (app) ratings are feedback provided voluntarily by users and serve as important evaluation criteria for apps. However, these ratings can often be biased owing to insufficient or missing votes. Additionally, significant differences have been observed between numeric ratings and user reviews. This study aims to predict the numeric ratings of Google apps using machine learning classifiers. It exploits numeric app ratings provided by users as training data and returns authentic mobile app ratings by analyzing user reviews. An ensemble learning model is proposed for this purpose that considers term frequency/inverse document frequency (TF/IDF) features. Three TF/IDF features, including unigrams, bigrams, and trigrams, were used. The dataset was scraped from the Google Play store, extracting data from 14 different app categories. Biased and unbiased user ratings were discriminated using TextBlob analysis to formulate the ground truth, from which the classifier prediction accuracy was then evaluated. The results demonstrate the high potential for machine learning-based classifiers to predict authentic numeric ratings based on actual user reviews.

• Asian-Australasian Journal of Animal Sciences
• /
• v.32 no.7
• /
• pp.1044-1051
• /
• 2019

Objective: Recent studies have implied that gene expression has high tissue-specificity, and therefore it is essential to investigate gene expression in a variety of tissues when performing the transcriptomic analysis. In addition, the gradual increase of long non-coding RNA (lncRNA) annotation database has increased the importance and proportion of mapped reads accordingly. Methods: We employed simple statistical models to detect the sexually biased/dimorphic genes and their conjugate lncRNAs in 40 RNA-seq samples across two factors: sex and tissue. We employed two quantification pipeline: mRNA annotation only and mRNA+lncRNA annotation. Results: As a result, the tissue-specific sexually dimorphic genes are affected by the addition of lncRNA annotation at a non-negligible level. In addition, many lncRNAs are expressed in a more tissue-specific fashion and with greater variation between tissues compared to protein-coding genes. Due to the genic region lncRNAs, the differentially expressed gene list changes, which results in certain sexually biased genes to become ambiguous across the tissues. Conclusion: In a past study, it has been reported that tissue-specific patterns can be seen throughout the differentially expressed genes between sexes in cattle. Using the same dataset, this study used a more recent reference, and the addition of conjugate lncRNA information, which revealed alterations of differentially expressed gene lists that result in an apparent distinction in the downstream analysis and interpretation. We firmly believe such misquantification of genic lncRNAs can be vital in both future and past studies.

• Convergence Security Journal
• /
• v.23 no.1
• /
• pp.27-33
• /
• 2023

Digitalization due to COVID-19 has rapidly developed artificial intelligence-based voice recognition technology. However, this technology causes unfair social problems, such as race and gender discrimination if datasets are biased against some groups, and degrades the reliability and security of artificial intelligence services. In this work, we compare and analyze accuracy-based unfairness in biased data environments using VGGNet (Visual Geometry Group Network), ResNet (Residual Neural Network), and MobileNet, which are representative CNN (Convolutional Neural Network) models of artificial intelligence. Experimental results show that ResNet34 showed the highest accuracy for women and men at 91% and 89.9%in Top1-accuracy, while ResNet18 showed the slightest accuracy difference between genders at 1.8%. The difference in accuracy between genders by model causes differences in service quality and unfair results between men and women when using the service.

• The Journal of Society for e-Business Studies
• /
• v.18 no.4
• /
• pp.315-326
• /
• 2013

Users in social media post various types of opinions such as product reviews and movie reviews. It is a common trend that customers get assistance from the opinions in making their decisions. However, as opinion usage grows, distorted feedbacks also have increased. For example, exaggerated positive opinions are posted for promoting target products. So are negative opinions which are far from common evaluations. Finding these biased opinions becomes important to keep social media reliable. Techniques of opinion mining (or sentiment analysis) have been developed to determine sentiment polarity of opinionated documents. These techniques can be utilized for finding the biased opinions. However, the previous techniques have some drawback. They categorize the text into only positive and negative, and they also need a large amount of training data to build the classifier. In this paper, we propose methods for discovering the biased opinions which are skewed from the overall common opinions. The methods are based on angle based outlier detection and personalized PageRank, which can be applied without training data. We analyze the performance of the proposed techniques by presenting experimental results on a movie review dataset.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.7 no.9
• /
• pp.2246-2267
• /
• 2013

BitTorrent is one of the most important applications in the current Internet. Despite of its interest, we still have little knowledge regarding the connectivity properties of real BitTorrent swarms. In this paper we leverage a dataset including the connectivity information of 250 real torrents and more than 150k peers to carefully study the connectivity properties of peers. The main topology parameters of the studied swarms suggest that they are significantly less resilient than random graphs. The analysis of the peer level connectivity properties reveals that peers continuously change more than half of their neighbours. Furthermore, we also find that a leecher typically keeps stable connections with a handful of neighbours with which it exchanges most of its traffic whereas seeders do not establish long-term connections with any peer so that they can homogeneously distribute chunks among leechers. Finally, we have discovered that a significant portion of the studied peers (45%) have an important locality-biased neighbourhood composition.

• Communications for Statistical Applications and Methods
• /
• v.28 no.2
• /
• pp.151-159
• /
• 2021

In this article, our interest is to estimate the association between consecutive gap times which are subject to interval censoring. Such data are referred as doubly interval censored data (Sun, 2006). In a context of serial event, an induced dependent censoring frequently occurs, resulting in biased estimates. In this study, our goal is to propose a Kendall's 𝜏 based association measure for doubly interval censored data. For adjusting the impact of induced dependent censoring, the inverse probability censoring weighting (IPCW) technique is implemented. Furthermore, a multiple imputation technique is applied to recover unknown failure times owing to interval censoring. Simulation studies demonstrate that the suggested association estimator performs well with moderate sample sizes. The proposed method is applied to a dataset of children's dental records.

• Journal of KIISE
• /
• v.43 no.2
• /
• pp.196-203
• /
• 2016

Due to the limitations of budget, resources, and time invested in a project, software requirements should be prioritized and be implemented in order of importance. Existing approaches to prioritizing requirements mostly depend on human decisions. The manual prioritization process is based on intensive interactions with the stakeholders, thus raising the issues of scalability and biased prioritization. To solve these problems, we propose a fully automated requirements prioritization approach, ToMSN (Topic Modeling Stakeholder Needs for requirements prioritization), by topic modeling the stakeholder needs-artifacts earned in the requirements elicitation phase. The requirements dataset of a 30,000-user system was utilized for the performance evaluation. ToMSN showed competitive prioritizing accuracy with existing approaches without human aids, therefore solving scalability and biased prioritization issues.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.8
• /
• pp.999-1005
• /
• 2020

While search portals' 'Portal News' account for the largest portion of aggregated news outlet, its neutrality as an outlet is questionable. This is because news aggregation may lead to prejudiced information consumption by recommending biased news articles. In this paper we introduce a new method of measuring political bias of news articles by using deep learning. It can provide its readers with insights on critical thinking. For this method, we build the dataset for deep learning by analyzing articles' bias from keywords, sourced from the National Assembly proceedings, and assigning bias to said keywords. Based on these data, news article bias is calculated by applying deep learning with a combination of Convolution Neural Network and Recurrent Neural Network. Using this method, 95.6% of sentences are correctly distinguished as either conservative or progressive-biased; on the entire article, the accuracy is 46.0%. This enables analyzing any articles' bias between conservative and progressive unlike previous methods that were limited on article subjects.

• Journal of the Korea Society for Simulation
• /
• v.33 no.2
• /
• pp.27-35
• /
• 2024

We aimed to shed light on the issue of perpetuating societal disparities by analyzing the impact of inherent biases present in datasets used for training artificial intelligence models on the predictions generated by Artificial Intelligence(AI). Therefore, to examine the influence of data bias on AI models, we constructed an original dataset containing biases related to gender wage gaps and subsequently created a de-identified dataset. Additionally, by utilizing the decision tree algorithm, we compared the outputs of AI models trained on both the original and de-identified datasets, aiming to analyze how data de-identification affects the biases in the results produced by artificial intelligence models. Through this, our goal was to highlight the significant role of data de-identification not only in safeguarding individual privacy but also in addressing biases within the data.