• The Journal of Information Systems
• /
• v.32 no.4
• /
• pp.119-136
• /
• 2023

Purpose We introduce a groundbreaking surveillance system explicitly designed to overcome the limitations typically associated with conventional surveillance systems, which often focus primarily on object-centric behavior analysis. Design/methodology/approach The study introduces an innovative approach to risk assessment in surveillance, employing image captioning to generate descriptive captions that effectively encapsulate the interactions among objects, actions, and spatial elements within observed scenes. To support our methodology, we developed a distinctive dataset comprising pairs of [image-caption-danger score] for training purposes. We fine-tuned the BLIP-2 model using this dataset and utilized BERT to decipher the semantic content of the generated captions for assessing risk levels. Findings In a series of experiments conducted with our self-constructed datasets, we illustrate that these datasets offer a wealth of information for risk assessment and display outstanding performance in this area. In comparison to models pre-trained on established datasets, our generated captions thoroughly encompass the necessary object attributes, behaviors, and spatial context crucial for the surveillance system. Additionally, they showcase adaptability to novel sentence structures, ensuring their versatility across a range of contexts.

• Korean Journal of Cognitive Science
• /
• v.14 no.2
• /
• pp.1-13
• /
• 2003

We introduce a way to generate morphologically and syntactically well-formed English sentences when building Korean to English bilingual dictionary for Machine Translation Systems. It has been proved that basic inflectional or structural descriptions for English sentences are by no means enough to generate proper English sentences because of traditional dictionary structures. Furthermore, much research has been focused only on how to disambiguate semantic ambiguities of words in a bilingual dictionary To take advantage of existing paperback Korean to English bilingual dictionary, its automatic conversion to an electronic version and methodologies to assign proper features to the descriptions for well-formed English sentences with minimum human effort have been proposed on the basis of the dictionary-specific structures. This approach was originally motivated for a glossary-based machine translation system, but it can be also applied to large scale dictionary work.

• Journal of KIISE:Software and Applications
• /
• v.35 no.2
• /
• pp.128-136
• /
• 2008

Recognition of parallel structure at early stage of sentence parsing can reduce the complexity of parsing. In this paper, we propose an unsupervised language-independent probabilistic model for recongition of parallel noun structures. The proposed model is based on the idea of swapping constituents, which replies the properties of symmetry (two or more identical constituents are repeated) and of reversibility (the order of constituents is inter-changeable) in parallel structures. The non-symmetric patterns that cannot be captured by the general symmetry rule are resolved additionally by the modifier information. In particular this paper shows how the proposed model is applied to recognize Korean parallel noun phrases connected by "wa/kwa" particle. Our model is compared with other models including supervised models and performs better on recongition of parallel noun phrases.

• The Journal of the Korea Contents Association
• /
• v.22 no.10
• /
• pp.74-82
• /
• 2022

The purpose of this paper is to provide a multi-modular account of double accusative constructions in Korean in the framework of Autolexical Grammar. The grammar views syntactic, semantic, and morphological structures of sentences as modules which are generated simultaneously and independently. Unlike syntactocentric theories, this paper analyzes semantic characteristics of double accusatives through function-argument (F/A) structure along with roles structure (RS) and information structure (IS). In F/A structure of double accusatives, the first accusative becomes an argument of a predicate, unlike the possessive, which is an argument of a relational noun. Furthermore, the first accusative of double accusatives takes the role of patient in RS, which allows it to become the subject of a passive sentence. On the other hand, the second accusative, which is originally the possessee, becomes a focal area in IS. Therefore, the purpose of double accusatives is twofold: one is to turn the possessor into an independent argument of a predicate which takes patient role, and the other is to turn the possessee into a focus. Such semantic characteristics of double accusatives can be expressed by means of multi-dimensional structures of F/A structure, RS, and IS of Autolexical Grammar, which allows an integrated account of the phenomenon.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.3
• /
• pp.359-366
• /
• 2022

In this paper, we present the Korean menu-ordering Sentence Text-to-Speech (TTS) system using conformer-based FastSpeech2. Conformer is the convolution-augmented transformer, which was originally proposed in Speech Recognition. Combining two different structures, the Conformer extracts better local and global features. It comprises two half Feed Forward module at the front and the end, sandwiching the Multi-Head Self-Attention module and Convolution module. We introduce the Conformer in Korean TTS, as we know it works well in Korean Speech Recognition. For comparison between transformer-based TTS model and Conformer-based one, we train FastSpeech2 and Conformer-based FastSpeech2. We collected a phoneme-balanced data set and used this for training our models. This corpus comprises not only general conversation, but also menu-ordering conversation consisting mainly of loanwords. This data set is the solution to the current Korean TTS model's degradation in loanwords. As a result of generating a synthesized sound using ParallelWave Gan, the Conformer-based FastSpeech2 achieved superior performance of MOS 4.04. We confirm that the model performance improved when the same structure was changed from transformer to Conformer in the Korean TTS.

• Journal of Industrial Convergence
• /
• v.18 no.4
• /
• pp.93-99
• /
• 2020

The Journal of Industrial Management Society in Republic of Korea. In order to explore the effect of overseas language training on the development of foreign language accuracy, this study investigates the errors in English compositions produced by 27 Korean university students who received overseas language training for 15 weeks. For data collection, students were made to take two tests, a pretest and a posttest, a semester apart. The differences in composition elements and errors between the two tests were examined and statistical analyses were performed. Results showed that while the average length of the compositions and sentences increased, the number of sentences decreased in the posttest. Also, more errors were found in the posttest where the students tried to construct more complex sentence structures. The students' ability to generate sentences were found to have improved, while their competence in using grammatical elements accurately within sentences did not see great improvement. This implies that overseas language training was not effective for aiding the development of one's grammatical accuracy of a foreign language over a 15-week period for the students.

• Journal of KIISE:Software and Applications
• /
• v.35 no.10
• /
• pp.627-635
• /
• 2008

An extended form of lexical network is explored by presenting U-WIN, which applies lexical relations that include not only semantic relations but also conceptual relations, morphological relations and syntactic relations, in a way different with existing lexical networks that have been centered around linking structures with semantic relations. So, This study introduces the new methodology for constructing a syntactic relation automatically. First of all, we extract probable nouns which related to verb based on verb's sentence type. However we should decided the extracted noun's meaning because extracted noun has many meanings. So in this study, we propose that noun's meaning is decided by the example matching rule/syntactic pattern/semantic similarity, frequency information. In addition, syntactic pattern is expanded using nouns which have high frequency in corpora.

• The Journal of the Korea Contents Association
• /
• v.10 no.5
• /
• pp.427-435
• /
• 2010

We extracted core terms by constructing scientific item networks from textbooks, analyzing their structures, and investigating the connected information and their relationships. For this research, we chose three high-school textbooks from different publishers for each three subjects, i.e, Science, Biology I and Biology II, to construct networks by linking scientific items in each sentence, where used items were regarded as nodes. Scientific item networks from all textbooks showed scare-free character. When core networks were established by applying k-core algorithm which is one of generally used methods for removing lesser weighted nodes and links from complex network, they showed the modular structure. Science textbooks formed four main modules of physics, chemistry, biology and earth science, while Biology I and Biology II textbooks revealed core networks composed of more detailed specific items in each field. These findings demonstrate the structural characteristics of networks in textbooks, and suggest core scientific items helpful for students' understanding of concept in Science and Biology.

• Speech Sciences
• /
• v.9 no.4
• /
• pp.187-200
• /
• 2002

The purpose of the present study is to see in terms of $F_{0}$ variation in Japanese how discourse focus and the lexical word accent interact with each other in realizing overall intonation patterns. Discourse focus causes prosodic restructuring of phrase structures and, as a result, largely affects pitch contours, whereas the lexical word accent is said to delimit the $F_{0}$ into a certain range. Measurement of $F_{0}$ was made of utterances of Japanese sentences to observe behavior of pitch contours with varied focus assignment and lexical accent specifications. The utterances were obtained in question-answer discourse contexts so that in a sentence, either one NP was always focused or no focus was assigned. I set four points for $F_{0}$ measurement; $F_{1s},F_{1m}, F_{2s}$, and $F_{2m}$, two for each noun phrase corresponding to $F_{0}$ at the beginning of the first syllable and that of the vocalic portion of the second syllable in the two NP's. The results of present study were as follows: (1) for all combination of lexical accent types, the $F_{0}$ rise both in NP1 and NP2 are higher when focused than when not focused. (2) NP2 starts a new accentual phrase when focused, showing even higher $F_{0}$ than NP1, the latter of which implies that in forming a new accentual phrase by focusing, catathesis does not seem to take effect on NP2 preceded by accented NP1. (3) unfocused NP2 preceded by unaccented NP1 has higher $F_{0}$ than those preceded by accented NP1.

• Journal of KIISE:Software and Applications
• /
• v.27 no.4
• /
• pp.448-462
• /
• 2000

Coordinate constructions in natural language pose a number of difficulties to natural language processing units, due to the increased complexity of syntactic analysis, the syntactic ambiguity of the involved lexical items, and the apparent deletion of predicates in various places. In this paper, we address the syntactic characteristics of the coordinate constructions in Korean from the viewpoint of constructing a competence grammar, and present a version of combinatory categorial grammar for the analysis of coordinate constructions in Korean. We also show how to utilize a unified lexicon in the proposed grammar formalism in deriving the sentential semantics and associated information structures as well, in order to capture the discourse functions of coordinate constructions in Korean. The presented analysis conforms to the common wisdom that coordinate constructions are utilized in language not simply to reduce multiple sentences to a single sentence, but also to convey the information of contrast. Finally, we provide an analysis of sample corpora for the frequency of coordinate constructions in Korean and discuss some problematic cases.