• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.6
• /
• pp.2470-2491
• /
• 2018

Web-scale open information extraction (Open IE) plays an important role in NLP tasks like acquiring common-sense knowledge, learning selectional preferences and automatic text understanding. A large number of Open IE approaches have been proposed in the last decade, and the majority of these approaches are based on supervised learning or dependency parsing. In this paper, we present a novel method for web scale open information extraction, which employs cosine distance based on Google word vector as the confidence score of the extraction. The proposed method is a purely unsupervised learning algorithm without requiring any hand-labeled training data or dependency parse features. We also present the mathematically rigorous proof for the new method with Bayes Inference and Artificial Neural Network theory. It turns out that the proposed algorithm is equivalent to Maximum Likelihood Estimation of the joint probability distribution over the elements of the candidate extraction. The proof itself also theoretically suggests a typical usage of word vector for other NLP tasks. Experiments show that the distance-based method leads to further improvements over the newly presented Open IE systems on three benchmark datasets, in terms of effectiveness and efficiency.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.3
• /
• pp.821-831
• /
• 1999

Recognition of three-dimensional objects with convexities and concavities is a hard and challenging problem. This paper presents a feature extraction method out of three-dimensional objects for the purpose of classification. This new method not only provides invariance to scale, translation, and rotation $R^3$ but also distinguishes any three-dimensional model objects with concavities and convexities by measuring a relative similarity in the information space where a set of characteristics features of objects is mapped.

• Journal of information and communication convergence engineering
• /
• v.20 no.2
• /
• pp.113-124
• /
• 2022

News in the form of web data generates increasingly large amounts of information as unstructured text. The capability of understanding the meaning of news is limited to humans; thus, it causes information overload. This hinders the effective use of embedded knowledge in such texts. Therefore, Automatic Knowledge Extraction (AKE) has now become an integral part of Semantic web and Natural Language Processing (NLP). Although recent literature shows that AKE has progressed, the results are still behind the expectations. This study proposes a method to auto-extract surface knowledge from English news into a machine-interpretable semantic format (triple). The proposed technique was designed using the grammatical structure of the sentence, and 11 original rules were discovered. The initial experiment extracted triples from the Sri Lankan news corpus, of which 83.5% were meaningful. The experiment was extended to the British Broadcasting Corporation (BBC) news dataset to prove its generic nature. This demonstrated a higher meaningful triple extraction rate of 92.6%. These results were validated using the inter-rater agreement method, which guaranteed the high reliability.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.31 no.4
• /
• pp.293-300
• /
• 2013

Road signs are important infrastructures for safe and smooth traffic by providing useful information to drivers. It is necessary to establish road sign DB for managing road signs systematically. To provide such DB, manually detection and recognition from imagery can be done. However, it is time and cost consuming. In this study, we proposed algorithms for automatic recognition of direction information in road sign image. Also we developed algorithm code using OpenCV library, and applied it to road sign image. To automatically detect and recognize direction information, we developed program which is composed of various modules such as image enhancement, image binarization, arrow region extraction, interesting point extraction, and template image matching. As a result, we can confirm the possibility of automatic recognition of direction information in road sign image.

• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06c
• /
• pp.258-261
• /
• 2011

Living in the age of information people face problems related to information overload. Information is easy to produce, store and distribute through various communication channels, one of which is emails. With the appearance of the mobile devices, such as smart phones and tabs, people can have access to email inbox at any moment of time from everywhere. In this paper we present information extraction system with a specific goal of extracting meeting location from the announcement of seminar or conference. We apply a machine learning method (conditional random fields, CRF), train the system using annotated corpus of seminar and conference announcements and validate results by applying various extracted correction rules and patterns. Furthermore, we normalize extracted location, and reference using geo-coding databases, OpenStreetMap and Wikipedia resources to determine real geographical coordinates.

• Journal of Korea Multimedia Society
• /
• v.18 no.10
• /
• pp.1172-1179
• /
• 2015

A wireless license plate region extracting module is proposed for LPR system controlling multiple gates. This module is cheaply implemented using Raspberry Pi which is open source and high performance. First, as the upper 1/3 of the captured image is discarded as it has no useful information on license plate. Using the OpenCV libraries the edge image is got by Canny algorithm after applying Gaussian filtering to gray image, and the labeling is conducted for 4 consecutive numbers in license plate. These numbers are located using various decision equations, and expanding the numbers region the final license plate region can be extracted. The result image is transferred to Server using wifi direct. Using the proposed module it becomes easy to set up and maintain the LPR system. The experimental results showed that the successful extracting rate was 98.4% using 500 car images with 640 × 480 resolution.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.41 no.12
• /
• pp.21-26
• /
• 2004

In this paper, a direct method is developed to extact RF equivalent circuit of a packaged BJT without optimization. First, parasitic components of plastic package are removed from measured S-parameters using open and short package patterns. Using package do-embedded S-parameters, a direct and simple method is proposed to extract bonding wire inductance and chip pad capacitance between package lead and chip pad. The small-signal model parameters of internal BJT are next determined by Z and Y-parameter formula derived from RF equivalent circuit. The modeled S-parameters of packaged BJT agree well with measured ones, verifying the accuracy of this new extraction method.

• Proceedings of the KSRS Conference
• /
• 2008.10a
• /
• pp.45-48
• /
• 2008

Since the mid-2000, so-called Web 2.0 paradigm was emerged and has been widely extended to actual Web markets with supporting Web technologies. Among Web 2.0 services, mashups with Open APIs is regarded as one of core technologies. Web mapping is also the most dominant mashup application in Open API-based applications, so that various mashups with Google Map API and other Open APIs are developing. In this paper, current technological status is explained with web GIS file formats of KML. While, a prototype is web-based GIS layer authoring system for general users, without any kinds of GIS tools and external spatial databases. This mashup application will be used web-based feature extraction for geo-spatial web for public users.

• Journal of Broadcast Engineering
• /
• v.19 no.3
• /
• pp.316-328
• /
• 2014

In this paper, we parallelize the popular feature detection algorithms, i.e. SIFT and SURF, and its application to fast panoramic image generation on the latest embedded GPU. Parallelized algorithms are implemented using recently developed OpenCL as the embedded GPGPU software platform. We compare the implementation efficiency and speed performance of conventional OpenGL Shading Language and OpenCL. Experimental result shows that implementation on OpenCL has comparable performance with GLSL. Compared with the performance on the embedded CPU in the same application processor, the embedded GPU runs 3~4 times faster. As an example of using feature extraction, panorama image synthesis is performed on embedded GPU by applying image matching using detected features.

• Proceedings of the KSRS Conference
• /
• 1999.11a
• /
• pp.310-315
• /
• 1999

We extracted narrow open water streams from JERS-1 SAR images of the Amazon rain forest. The extracted range of these streams were almost comparable to a high level extraction of the same streams from near-IR images of JERS-1 VNIR data notwithstanding that these features in SAR images show the strong dependence of the observation angle. Large water bodies are relatively easy to extract from JERS-1 SAR images, as they tend to appear as very dark areas; but streams whose width is nearly equal to or less than the spatial resolution no longer appear as very dark features. By using strong scatterers distributed sparsely along the radar facing sides of the streams, we can successfully estimate approximate ranges of waterways and then extract relatively dark line-like features within these ranges.