• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.5
• /
• pp.1413-1432
• /
• 2023

Recommender systems aim to recommend items to users by taking into account their probable interests. This study focuses on creating a model that utilizes multiple sources of information about users and items by employing a multimodality approach. The study addresses the task of how to gather information from different sources (modalities) and transform them into a uniform format, resulting in a multi-modal feature description for users and items. This work also aims to transform and represent the features extracted from different modalities so that the information is in a compatible format for integration and contains important, useful information for the prediction model. To achieve this goal, we propose a novel multi-modal recommendation model, which involves extracting latent features of users and items from a utility matrix using matrix factorization techniques. Various transformation techniques are utilized to extract features from other sources of information such as user reviews, item descriptions, and item categories. We also proposed the use of Principal Component Analysis (PCA) and Feature Selection techniques to reduce the data dimension and extract important features as well as remove noisy features to increase the accuracy of the model. We conducted several different experimental models based on different subsets of modalities on the MovieLens and Amazon sub-category datasets. According to the experimental results, the proposed model significantly enhances the accuracy of recommendations when compared to SVD, which is acknowledged as one of the most effective models for recommender systems. Specifically, the proposed model reduces the RMSE by a range of 4.8% to 21.43% and increases the Precision by a range of 2.07% to 26.49% for the Amazon datasets. Similarly, for the MovieLens dataset, the proposed model reduces the RMSE by 45.61% and increases the Precision by 14.06%. Additionally, the experimental results on both datasets demonstrate that combining information from multiple modalities in the proposed model leads to superior outcomes compared to relying on a single type of information.

• Educational Technology International
• /
• v.18 no.1
• /
• pp.49-71
• /
• 2017

The purpose of the study was to identify key design features of virtual reality with head-mounted displays (HMD) and touchless interface for the hearing-impaired and hard-of-hearing learners. The virtual reality based training program was aimed to help hearing-impaired learners in machine operating learning, which requires spatial understanding to operate. We developed an immersive virtual learning environment prototype with an HMD (Oculus Rift) and a touchless natural user interface (Leap Motion) to identify the key design features required to enhance virtual reality for the hearing-impaired and hard-of-hearing learners. Two usability tests of the prototype were conducted, which revealed that several features in the system need revision and that the technology presents an enormous potential to help hearing-impaired learners by providing realistic and immersive learning experiences. After the usability tests of hearing-impaired students' exploring the 3D virtual space, interviews were conducted, which also established that further revision of the system is needed, which would take into account the learners' physical as well as cognitive characteristics.

• ETRI Journal
• /
• v.46 no.2
• /
• pp.250-262
• /
• 2024

Air-writing recognition is relevant in areas such as natural human-computer interaction, augmented reality, and virtual reality. A trajectory is the most natural way to represent air writing. We analyze the recognition accuracy of words written in air considering five features, namely, writing direction, curvature, trajectory, orthocenter, and ellipsoid, as well as different parameters of a hidden Markov model classifier. Experiments were performed on two representative datasets, whose sample trajectories were collected using a Leap Motion Controller from a fingertip performing air writing. Dataset D₁ contains 840 English words from 21 classes, and dataset D₂ contains 1600 English words from 40 classes. A genetic algorithm was combined with a hidden Markov model classifier to obtain the best subset of features. Combination ftrajectory, orthocenter, writing direction, curvatureg provided the best feature set, achieving recognition accuracies on datasets D₁ and D₂ of 98.81% and 83.58%, respectively.

• Korean Institute of Interior Design Journal
• /
• v.23 no.6
• /
• pp.14-21
• /
• 2014

Kengo Kuma is one of the features of the building is 'natural architecture'. He is saying that the construction of the building in which he is seeking is a natural material of construction is only applied to the issue as to whether the material nature, but that the relationship between humans, the environment, natural point of view in his book. Like repeating units are part of the two hundred ninety-four came mainly small ways to create a relationship, or have architectural features make up the sum of his architecture. I present a comprehensive study is reported in terms of that the 'grain' of such a small part of the unit, according to the architecture of Kengo Kuma mainly granulation characteristics of the construction of one of the practical expression studies Kengo Kuma. Is represented as louvers, check pattern, hole, granulation unit such trend, but have become developed in other directions such simple types are combined. Analysis and particle characteristics in terms of the architectural space of Kengo Kuma, serve as the purpose of the present study reveal that each of the features that have a meaning in the construction of any Kengo Kuma.

• Journal of the Institute of Convergence Signal Processing
• /
• v.11 no.2
• /
• pp.143-150
• /
• 2010

Markerless augmented reality system must have the capability to recognize and match natural objects both in indoor and outdoor environment. In this paper, a novel approach is proposed for extracting features and recognizing natural objects using visual descriptors and codebooks. Since the augmented reality applications are sensitive to speed of operation and real time performance, our work mainly focused on recognition of multi-class natural objects and reduce the computing time for classification and feature extraction. SIFT(scale invariant feature transforms) and SURF(speeded up robust feature) are used to extract features from natural objects during training and testing, and their performance is compared. Then we form visual codebook from the high dimensional feature vectors using clustering algorithm and recognize the objects using naive Bayes classifier.

• Korean Institute of Interior Design Journal
• /
• v.20 no.1
• /
• pp.71-79
• /
• 2011

Mount Angel Abbey library shows not only the general characteristics which Aalto had shown in his 60s but also special features of his last library project. Special features of Mount Angel library are the result of considering geography and existing context of the site. General characteristics of the library mainly come from the Aalto's concept for library especially considering light. Due to steep site, Mount Angel Abbey library has multi-layered and dynamic spatial configuration. Aalto considered the importance of daylight in the library and applied various kinds of openings to induce natural light. Therefore, Mount Angel Abbey library is the final result of his library type that developed from the Viipuri library of 1930s and his architectural thoughts of 1960s. As a combination of space and light, Mount Angel Abbey library is the outcome of Aalto's unique functional thoughts which maximize the function of the library through efficient induction of natural light.

• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.978-980
• /
• 2003

This study explores the temporal and spatial features near 67years (1934 ?2001) and landcover change in last 14 years (1987-2001) in Shijiazhuang, China, based on 67-year time series data edited from historical maps, TM and ETM+ imageries by integrating GIS and remote sensing method. An index named Annual Growth Rate (AGR) is used to analyze the spatial features of urban sprawl, and Maximum Likelihood classification method is utilized to detect the land cover types change. At last, the relationship between urbanization and factors is analyzed.

• The Bulletin of The Korean Astronomical Society
• /
• v.22
• /
• pp.6-6
• /
• 1997

• KIEAE Journal
• /
• v.14 no.1
• /
• pp.5-13
• /
• 2014

The ancestors of China created various proper dwellings in the vast Chinese lands, according to distinct geographical environment, climate, and special local products. The purpose of this study is to compare features of 7 typical traditional Chinese dwellings in different areas from historical, geographical, structural and tinctorial aspects, in order to see what the differences are and rearrange them by these 4 different scales as estimate standards. By Comparing features between 7 traditional dwellings, advantages and disadvantages of them were found out. Not only natural terrain but also cultural mentality is an important factor that effected on the changes of regionality. The result showed that historical length in a certain extent reflected the differences between dwellings in different areas; there are more regular type dwellings in the northern China; dwellings in different structural material types following their regions; and, southern dwellings prefer natural and simple exterior colors. Also From construction aesthetics and culture perspective, Different cultures have cultivated and brought out different dwelling styles. Thus, protecting on traditional dwellings is of great necessities and this study will bring benefits in many aspects. For a further research, the result of this study can be utilized as the base data that suggest directions for effective Chinese residence planning in different areas meeting different demands.

• Bulletin of the Korean Chemical Society
• /
• v.35 no.4
• /
• pp.1244-1246
• /
• 2014