• Proceedings of the KIEE Conference
• /
• 1987.07b
• /
• pp.1046-1049
• /
• 1987

This paper proposes a Korean-Japanese Machine Translation System that processes unregistered words, modifiers and compound sentences. In mophological analysis, the unregistered words are processed by using unregistered word processing algorithm. The modifiers are processed by consulting noun-attributes and grammar rules. The compound sentence processing algorithm recognizes whether the sentence that includes commas is compound sentence or not. This system performs on IBM-PC/AT DOS using Prolog-1.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.11
• /
• pp.124-135
• /
• 1994

In this paper, we propose a Korean syntax analysis system for special syntax processing. HPSG, which processes syntatic and semantic analysis unificationally, is chosen for grammar description. Head-driven unidirectional active chart parser, which is efficient in Korean processing, is used for parsing mechanism. The parser of this paper can analyze not only general sentence structure which consists of complement-head, adjunct-head and head-head structure bur also special syntax which consists of auxiliay verb sentence, causative sentence, passive sentence and so on.

• Journal of Institute of Control, Robotics and Systems
• /
• v.15 no.1
• /
• pp.111-119
• /
• 2009

For daily life interaction with human, robots need the capability of encoding and storing cognitive information and retrieving it contextually. In this paper, spatiotemporal grounding of cognitive information for a language based cognitive system is presented. The cognitive information of the event occurred at a robot is described with a sentence, stored in a memory, and retrieved contextually. Each sentence is parsed, discriminated with the functional type of it, and analyzed with argument structure for connecting to cognitive information. With the proposed grounding, the cognitive information is encoded to sentence form and stored in sentence memory with object descriptor. Sentences are retrieved for answering questions of human by searching temporal information from the sentence memory and doing spatial reasoning in schematic imagery. An experiment shows the feasibility and efficiency of the spatiotemporal grounding for advanced service robot.

• The KIPS Transactions:PartB
• /
• v.12B no.3 s.99
• /
• pp.323-328
• /
• 2005

This study is sentence generation for Augmentative and Alternative Communication. The object of sentence generation is to use in augmentative and alternative communication which is designed for those who are nonspeaking disorders. AAC generates human voice with using a sentence which is made up by the users. In order to construct a sentence, lexical information was adapted for a concept of augmentative and alternative communication. The lexical informations consist of noun types which can be connected to verbs, auxiliary words, conjugation of verbs and verb types. The system was made using lexical information and the usefulness of the sentence generation was measured by the system. The system constructed has functions of generation and saving right sentences, searching and inputting vocabularies.

• The KIPS Transactions:PartB
• /
• v.11B no.5
• /
• pp.563-572
• /
• 2004

Sentence ion is a simplification of a sentence preserving its communicative function. It accomplishes sentence revision and concept ion simultaneously. Sentence revision is a method that resolves the discrepancy between human's thoughts and its expressed semantic in sentences. Concept ion is an expression of general ideas acquired from the common elements of concepts. Sentence ion selects the main constituents of given sentences and describes the upper concepts of them with detecting their semantic information. This enables sen fence revision and concept ion simultaneously. In this paper, a syntactic parser LGPI+ and an ontology OfN are utilized for sentence ion. Sentence abstracter SABOT makes use of LGPI+ and OfN. SABOT processes the result of parsing and selects the candidate words for sentence ion. This paper computes the sentence recall of the main sentences and the topic hit ratio of the selected sentences with the text understanding system using sentence ion. The sources are 58 paragraphs in 23 stories. As a result of it, the sentence recall is about .54 ～ 72% and the topic hit ratio is about 76 ～ 86%. This paper verified that sentence ion enables sentence revision that can select the topic sentences of a given text efficiently and concept ion that can improve the depth of text understanding.

• Proceedings of the IEEK Conference
• /
• 2003.07e
• /
• pp.2419-2422
• /
• 2003

To improve performance of sentence speech recognition systems, we need to consider perplexity of language model and the number of words of dictionary for increasing vocabulary size. In this paper, we propose a language model of VCCV units for sentence speech recognition. For this, we choose VCCV units as a processing units of language model and compare it with clauses and morphemes. Clauses and morphemes have many vocabulary and high perplexity. But VCCV units have small lexicon size and limited vocabulary. An advantage of VCCV units is low perplexity. This paper made language model using bigram about given text. We calculated perplexity of each language processing unit. The perplexity of VCCV units is lower than morpheme and clause.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.05a
• /
• pp.640-643
• /
• 2008

현재 우리나라는 좁은 국토와 도시 집중화로 인해 급증하고 있는 지하매설물의 체계적인 관리가 어느 때보다 요구되고 있다. 그러나 다른 분야에서 빠르게 진행되고 있는 전산화 및 정보화 작업에 비해 상대적으로 더딘 지리정보체계 구축으로 지하매설물 관리를 위한 정보시스템이 미비한 상태이다. 이는 지리정보체계 구축 방법에 어려움과 소요되는 인력, 비용, 시간 등이 크기 때문으로 판단된다. 본 논문에서는 GPS와 능동형 RFID를 이용한 위치기반 지하매설물 정보시스템의 구현을 기술하고 시뮬레이션하였다. 이 시스템은 GPS NMEA Sentence의 위치정보와 능동형 RFID의 인식정보를 결합한 데이터를 이용하여 정확하고 체계적인 지하매설물 정보 구축이 가능하고, 효율적인 지하매설물 관리를 위한 인터페이스를 제공한다.

• Journal of the Institute of Convergence Signal Processing
• /
• v.5 no.3
• /
• pp.198-204
• /
• 2004

This study suggests a sentence generating method to help those who have language impediment with their communication. The method suggested in this study was constructed into a system in order to be applied to AAC system. AAC system is a personal portable device that generates sentences. Those who have language impediment need another communication method, causes inconvenience when used in a conversation with those who don't have the same trouble. The method of inputting both consonants and vowels can be inconvenient and time consuming for a conversational communication because of the number of the key strokes. The lexical information for the sentence generating of this study defines the user's domain, collects the adequate words and sentences, and extracts and classifies the characteristics of the collected words. The comparison between the number of key strokes for sentence generating using the system and that of inputting consonants and vowels using a keyboard was made in order to evaluate the usefulness the sentence generating method.

• Journal of Korean Society for Quality Management
• /
• v.48 no.3
• /
• pp.421-431
• /
• 2020

Purpose: The purpose of this study is to compare and analyze the tokenizer in natural language processing for customer satisfaction in sentiment analysis. Methods: In this study, a supervised learning-based tokenizer Mecab-Ko and an unsupervised learning-based tokenizer SentencePiece were used for comparison. Three algorithms: Naïve Bayes, k-Nearest Neighbor, and Decision Tree were selected to compare the performance of each tokenizer. For performance comparison, three metrics: accuracy, precision, and recall were used in the study. Results: The results of this study are as follows; Through performance evaluation and verification, it was confirmed that SentencePiece shows better classification performance than Mecab-Ko. In order to confirm the robustness of the derived results, independent t-tests were conducted on the evaluation results for the two types of the tokenizer. As a result of the study, it was confirmed that the classification performance of the SentencePiece tokenizer was high in the k-Nearest Neighbor and Decision Tree algorithms. In addition, the Decision Tree showed slightly higher accuracy among the three classification algorithms. Conclusion: The SentencePiece tokenizer can be used to classify and interpret customer sentiment based on online reviews in Korean more accurately. In addition, it seems that it is possible to give a specific meaning to a short word or a jargon, which is often used by users when evaluating products but is not defined in advance.

• The KIPS Transactions:PartB
• /
• v.14B no.7
• /
• pp.559-566
• /
• 2007

Subjective sentences express opinions, emotions, evaluations and other subjective ideas relevant to products or events. These expressions sometimes can be seen in only part of a sentence, thus extracting features from a full-sentence can degrade the performance of subjective-sentence-classification. This paper presents a method for improving the performance of a subjectivity classifier by combining two classifiers generated from the different representations of an input sentence. One representation is a sentimental phrase that represents an automatically identified subjective expression or objective expression and the other representation is a full-sentence. Each representation is used to extract modified n-grams that are composed of a word and its contextual words' polarity information. The best performance, 79.7% accuracy, 2.5% improvement, was obtained when the phrase-level classifier and the sentence-level classifier were merged.