• Phonetics and Speech Sciences
• /
• v.15 no.3
• /
• pp.45-52
• /
• 2023

Previous studies have established a role for cognitive differences in explaining variability in speech processing across individuals. In the case of perceptual cue weighting in the context of a sound change, studies have produced conflicting results regarding the relationship between executive function and the use of redundant cues. The current study aimed to explore this relationship in acoustic cue weighting during speech production. Forty-one Korean-speaking children read a list of stop-initial words and completed two tests that assess executive function, i.e., Dimensional Change Card Sorting (DCCS) and digit n-back. Voice onset time (VOT) and fundamental frequency (F0) were measured in each word, and analyses were carried out to determine the extent to which children's executive function predicted their use of both informative and less informative cues to the three pairs comprising the Korean three-way stop laryngeal contrast. No evidence was found for a relationship between cognitive ability and acoustic cue weighting in production, which is at odds with previous, albeit conflicting, results for speech perception. While this result may be due to the lack of task demands in the production task used here, it nevertheless expands the empirical ground upon which future work in this area may proceed.

• MALSORI
• /
• no.60
• /
• pp.165-180
• /
• 2006

In this paper, we present several confidence measures (CM) for speech recognition systems to evaluate the reliability of recognition results. We propose heuristic CMs such as mean log-likelihood score, N-best word log-likelihood ratio, likelihood sequence fluctuation and likelihood ratio testing(LRT)-based CMs using several types of anti-models. Furthermore, we propose new algorithms to add weighting terms on phone-level log-likelihood ratio to merge word-level log-likelihood ratios. These weighting terms are computed from the distance between acoustic models and knowledge-based phoneme classifications. LRT-based CMs show better performance than heuristic CMs excessively, and LRT-based CMs using phonetic information show that the relative reduction in equal error rate ranges between $8{\sim}13%$ compared to the baseline LRT-based CMs. We use the support vector machine to fuse several CMs and improve the performance of utterance verification. From our experiments, we know that selection of CMs with low correlation is more effective than CMs with high correlation.

• Journal of Advanced Marine Engineering and Technology
• /
• v.38 no.4
• /
• pp.444-450
• /
• 2014

Most researches on word sense disambiguation have used static sized context regardless of sentence patterns. This paper proposes to use dynamic sized context considering sentence patterns and distance between words for word sense disambiguation. We evaluated our system 12 words in 32,735sentences with Sejong POS and sense tagged corpus, and dynamic sized context showed 92.2% average accuracy for predicates, which is better than accuracy of static sized context.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.1 s.39
• /
• pp.273-282
• /
• 2006

In this paper, we propose index method introducing distance concept in word by a method weighting word. This index method is frequent representing an inquiry word and document index and compound noun or more than two adjoin nouns or noun phrase, the farther the distance between these nouns, the fewer selected ratio decreases in index point is the aiming, this choose guide word candidate by existent weight grant method and distance between candidates chose candidate finally in index within 3 sentences. Using in these way I document of 100 kinds of newspaper, scientific treatise, web document and so on, showed the correctness rate resulted of newspaper 92.03% scientific treatise 95% web document 73.33%.

• The Journal of Korean Academy of Prosthodontics
• /
• v.25 no.1
• /
• pp.71-94
• /
• 1987

This study was carried out to investigate the sound distortion following the alternation of the palatal plate thickness. For this study, 2 healthy male subjects (24-year-old) were selected. Born in Seoul, they both spoke Seoul dialect. First, their sounds of /na(나)/, /da(다)/, /1a(라)/, /ja(자)/, /cha(차)/, /ta(타)/, without inserting plates were recorded, and then the sounds with palatal plates of different thickness were recorded, successively. The plate was fabricated in 3 types, each palatal thickness being 1.0mm, 2.5mm, dentoalveolar portion 2.5mm, other residual portion was 1.0mm, successively. Each type plates named B, C, D-type, in succession. Series of analysis were administered through Computer(16 bit) to analyze the sound distortions. These experiments were analyzed by the LPC (without weighting, pre-weighting, post-weighting) of the consonants, vowels portion, formant frequency of the vowels and word duration of the consonants. The findings led to the following conclusions: 1. There was no correlation of the distortion rate on the 2 informants. 2. Generally, vowels were not affected by the palatal plate thickness in the formant analysis, however, more distortion was detected in the LPC analysis, especially C, D-type plates. 3. Consonants distortion was more evident in the C, D-type plate. 4. The second formant was most disturbed and reduced in the all consonants with insertion of the palatal plate, especially C, D-type plate. 5. Word duration was shortened in the plate inserted(except /ja/, /cha/), especially C, D-type. 6. It was found that dentoalveolar, hard palatal sounds were severely distorted in plate inserted, and they were mainly affected by the dentoalveolar portion thickness. 7. There was correlation between palatal thickness and consonants quality.

• The KIPS Transactions:PartB
• /
• v.11B no.3
• /
• pp.369-380
• /
• 2004

In this paper, we propose a web page weighting scheme based on WordNet-based collaborative evaluation and hyperlink to improve the precision of web search engine. Generally search engines use keyword matching to decide web page ranking. In the information retrieval from huge data such as the Web, simple word comparison cannot distinguish important documents because there exist too many documents with similar relevancy. In this paper, we implement a WordNet-based user interface that helps to distinguish different senses of query word, and constructed a search engine in which the implicit evaluations by multiple users are reflected in ranking by accumulating the number of clicks. In accumulating click counts, they are stored separately according to lenses, so that more accurate search is possible. Weighting of each web page by using collaborative evaluation and hyperlink is reflected in ranking. The experimental results with several keywords show that the precision of proposed system is improved compared to conventional search engines.

• Journal of KIISE:Software and Applications
• /
• v.31 no.8
• /
• pp.1101-1112
• /
• 2004

In this paper, we propose a Web page weighting scheme using the user preference in each sense of query word to improve the performance of Web search. Generally search engines assign weights to a web page by using relevancy only, which is obtained by comparing the query word and the words in a web page. In the information retrieval from huge data such as the Web, simple word comparison cannot distinguish important documents because there exist too many documents with similar relevancy In this paper we implement a WordNet-based user interface that helps to distinguish different senses of query word, and constructed a search engine in which the implicit evaluations by multiple users are reflected in ranking by accumulating the number of clicks. In accumulating click counts, they are stored separately according to senses, so that more accurate search is possible. The experimental results with several keywords show that the precision of proposed system is improved compared to conventional search engines.

• Journal of Practical Engineering Education
• /
• v.15 no.2
• /
• pp.509-517
• /
• 2023

The reactive web service provides a related word recommendation system using the TextRank algorithm and a word-based idea generation service selected by the user. In the related word recommendation system, the method of weighting each word using the TextRank algorithm and the probability output method using SoftMax are discussed. The idea generation service discusses the idea generation method and the artificial intelligence reinforce-learning method using mini-GPT. The reactive web discusses the linkage process between React, Spring Boot, and Flask, and describes the overall operation method. When the user enters the desired topic, it provides the associated word. The user constructs a mind map by selecting a related word or adding a desired word. When a user selects a word to combine from a constructed mind-map, it provides newly generated ideas and related patents. This web service can share generated ideas with other users, and improves artificial intelligence by receiving user feedback as a horoscope.

• Journal of the Korean Society for information Management
• /
• v.20 no.4 s.50
• /
• pp.233-248
• /
• 2003

The Inverse Document Frequency (IDF) weighting method is based on the hypothesis that in the document collection the lower the frequency of a term is, the more important the term is as a subject word. This well-known hypothesis is, however, somewhat questionable because some low frequency terms turn out to be insufficient subject words. This study suggests the pivoted IDF weighting method for better retrieval effectiveness, on the assumption that medium frequency terms are more important than low frequency terms. We thoroughly evaluated this method on three test collections and it showed performance improvements especially at high ranks.

• Korean Journal of Cognitive Science
• /
• v.13 no.1
• /
• pp.1-11
• /
• 2002

In this paper, a test was performed to determine whether words in Anchor Text were appropriate as key words. As a result of the test. there were proper words of high weighting factor, while some others did not even appear in the text. therefore, were not appropriate as key words. In order to resolve this problem. a new method was proposed to extract key words. Using the proposed method, inappropriate key words can be removed so that new key words be set, and then, ranking becomes possible with the TFIDF value as a weighting factor of the key word. It was verified that the new method has higher accuracy compared to the previous methods.