신동혁 교수

This paper proposes a fast noise estimation algorithm using a Gaussian filter. It is based on block-based noise estimation, in which an input image is assumed to be contaminated by the additive white Gaussian noise and a filtering process is performed by an adaptive Gaussian filter. Coefficients of a Gaussian filter are selected as functions of the standard deviation of the Gaussian noise that is estimated from an input noisy image. For estimation of the amount of noise (i.e., standard deviation

This research methods article proposes a visual data analytics framework to enhance social media research using deep learning models. Drawing on the literature of information systems and marketing, complemented with data-driven methods, we propose a number of visual and textual content features including complexity, similarity, and consistency measures that can play important roles in the persuasiveness of social media content. We then employ state-of-the-art machine learning approaches such as

The application a smart phone user will launch next intuitively depends on the sequence of apps used recently. More generally, when users interact with systems such as shopping websites or online radio, they click on items that are of interest in the current context. We call the sequence of clicks made in the current session interactional context. It is desirable for a recommender system to use the context set by the user to update recommendations. Most current context-aware recommender systems

Popular microblogging sites such as Tumblr have attracted hundreds of millions of users as a content sharing platform, where users can create rich content in the form of posts that are shared with other users who follow them. Due to the sheer amount of posts created on such services, an important task is to make quality recommendations of blogs for users to follow. Apart from traditional recommender system settings where the follower graph is the main data source, additional side-information of

Low-density parity-check (LDPC) codes have an inherent stopping criterion, parity-check constraints (equations). By testing the parity-check constraints, an LDPC decoder can detect successful decoding and stop their decoding, which is, however, not possible with turbo codes. In this paper, we propose a stopping criterion to predict decoding failure of LDPC codes, instead of detecting successful decoding. If the decoder predicts the decoding failure in advance, the receiver can more rapidly respo

The automated analysis of social networks has become an important problem due to the proliferation of social networks, such as LiveJournal, Flickr and Facebook. The scale of these social networks is massive and continues to grow rapidly. An important problem in social network analysis is proximity estimation that infers the closeness of different users. Link prediction, in turn, is an important application of proximity estimation. However, many methods for computing proximity measures have high

In the field of machine learning and pattern recognition, feature subset selection is an important area, where many approaches have been proposed. In this paper, we choose some feature selection algorithms and analyze their performance using various datasets from public domain. We measured the number of reduced features and the improvement of learning performance with chosen feature selection methods, then evaluated and compared each method on the basis of these measurements.

Drawing on the notion of compensatory behavior, this paper studies how students compensate for learning loss during a pandemic and what role artificial intelligence (AI) plays in this regard. We further probe into a difference in compensatory behavior for learning loss in terms of quantity, pattern, and pace (i.e., tripartite aspect of learning behavior) of AI-powered learning app usage depending on the level of pandemic threat and the proximity of a goal to students. Results show that the pande

Computing the k dominant eigenvalues and eigenvectors of massive graphs is a key operation in numerous machine learning applications; however, popular solvers suffer from slow convergence, especially when k is reasonably large. In this paper, we propose and analyze a novel multi-scale spectral decomposi-tion method (MSEIGS), which first clusters the graph into smaller clusters whose spectral decomposition can be computed efficiently and independently. We show theoretically as well as empirically

In microblogging sites, recommending blogs (users) to follow is one of the core tasks for enhancing user experience. In this paper, we propose a novel inductive matrix completion based blog recommendation method to effectively utilize multiple rich sources of evidence such as the social network and the content as well as the activity data from users and blogs. Experiments on a large-scale real-world dataset from Tumblr show the effectiveness of the proposed blog recommendation method.

Disinformation activities that aim to manipulate public opinion pose serious challenges to managing online platforms. One of the most widely used disinformation techniques is bot-assisted fake social engagement, which is used to falsely and quickly amplify the salience of information at scale. Based on agenda-setting theory, we hypothesize that bot-assisted fake social engagement boosts public attention in the manner intended by the manipulator. Leveraging a proven case of bot-assisted fake soci

We propose a new multi-class classification algorithm based on one-class SVM and nearest mean classifier methods. A wrapper-style feature selection scheme designed specifically for our algorithm is also provided for increased classification accuracy. It will be demonstrated that the proposed classification algorithm provide excellent performance, and in particular, performs strictly better than some of the currently known best classification algorithms on five biological datasets.

The automated analysis of social networks has become an important problem due to the proliferation of social networks, such as LiveJournal, Flickr and Facebook. The scale of these social networks is massive and continues to grow rapidly. An important problem in social network analysis is proximity estimation that infers the closeness of different users. Link prediction, in turn, is an important application of proximity estimation. However, many methods for computing proximity measures have high