• Journal of KIISE:Software and Applications
• /
• v.35 no.4
• /
• pp.245-254
• /
• 2008

This paper proposes a unified multi-modal input framework to interface the recognition engines such as IBM ViaVoice and Microsoft handwriting-recognition system with general window applications, particularly, for pen-input displays. As soon as user pushes a hardware button attached to the pin-input display with one hand, the current window of focus such as a internet search window and a word processor is overlaid with a transparent window covering the whole desktop; upon which user inputs handwriting with the other hand, without losing the focus of attention on working context. As well as freeform handwriting on this transparent input overlay as a sketch pad, the user can dictate some words and draw diagrams to communicate with the system.

• The KIPS Transactions:PartB
• /
• v.15B no.2
• /
• pp.171-178
• /
• 2008

Foreign word transliterations are not consistently used in documents, which hinders retrieving some important relevant documents in exact term matching information retrieval systems. In this paper, a meta search method is proposed, which expands and refines relevant variant queries from an original input foreign word transliteration query to retrieve the more relevant documents. The method firstly expands a transliteration query to the variants using a statistical method. Secondly the method selects the valid variants: it queries each variant to the retrieval systems beforehand and checks the validity of each variant by counting the number of appearance of the variant in the retrieved document and calculating the similarity of the context of the variant. Experiment result showed that querying with the variants produced at the first step, which is a base method of the test, performed 38% in average F measure, and querying with the refined variants at the second step, which is a proposed method, significantly improved the performance to 81% in average F measure.

• Journal of KIISE:Software and Applications
• /
• v.31 no.9
• /
• pp.1171-1179
• /
• 2004

We describe word boundary detection that extracts the boundary between speech and non-speech. The proposed method uses two features. One is the normalized root mean square of speech signal, which is insensitive to white noises and represents temporal information. The other is the normalized met-frequency band energy of voice signal, which is frequency information of the signal. Our method detects word boundaries using a recurrent fuzzy associative memory(RFAM) that extends FAM by adding recurrent nodes. Hebbian learning method is employed to establish the degree of association between an input and output. An error back-propagation algorithm is used for teaming the weights between the consequent layer and the recurrent layer. To confirm the effectiveness, we applied the suggested system to voice data obtained from KAIST.

• Proceedings of the Korea Contents Association Conference
• /
• 2006.05a
• /
• pp.140-143
• /
• 2006

In this paper, we describe a question-answering system which extracts answers for the superlative questions which include the superlative words such as "the most", "the best", "the first", "the largest", "the least", and so on. The superlative questions are composed of four main components and others. Four main components are the superlative word, answer type, regional information, and a verb modified by the superlative word. We classify the superlative words into two types as to whether the verb has to be needed to be a question or not. The superlative word, answer type and regional information are essential elements to extract answer for all superlative questions. But the verb may be an essential element by the type of superlative word. Our system analyzes input question, and finds four main components of the superlative question. Also, our system searches relative documents and candidate sentences using them, and extracts answers from candidate sentences. Empirical result shows that our system has high precision and high recall for the superlative questions.

• Journal of Digital Contents Society
• /
• v.18 no.7
• /
• pp.1419-1425
• /
• 2017

The ultimate goal of legal knowledge search is to obtain optimal legal information based on laws and precedent. Text mining research is actively being undertaken to meet the needs of efficient retrieval from large scale data. A typical method is to use a word embedding algorithm based on Neural Net. This paper demonstrates how to search relevant information, applying Korean law information to word embedding. First, we extracts reference laws from precedents in order and takes reference laws as input of Law2Vec. The model learns a law by predicting its surrounding context law. The algorithm then moves over each law in the corpus and repeats the training step. After the training finished, we could infer the relationship between the laws via the embedding method. The search performance was evaluated based on precision and the recall rate which are computed from how closely the results are associated to the search terms. The test result proved that what this paper proposes is much more useful compared to existing systems utilizing only keyword search when it comes to extracting related laws.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.19 no.9
• /
• pp.1702-1709
• /
• 1994

In the postprocessing of Hangul recognition system, the storage structure of contextual information is an important matter for the recognition rate and speed of the entire system. Trie in general is used to represent the context as word dictionary, but the memory space efficiency of the structure is low. Therefore we propose a new structure for word dictionary that has better space efficiency and the equivalent merits of trie. Because Hangul is a compound language, the language can be represented by phonemes or by characters. In the representation by phonemes(P-mode) the retrieval is fast, but the space efficiency is low. In the representation by characters(C-mode) the space efficiency is high, but the retrieval is slow. In this paper the two representation methods are combined to form a hybrid representation(H-mode). At first an optimal level for the combination is selected by two characteristic curves of node utilization and dispersion. Then the input words are represented with trie structure by P-mode from the first to the optimal level, and the rest are represented with sequentially linked list structure by C-mode. The experimental results for the six kinds of word set show that the proposed structure is more efficient. This result is based on the fact that the retrieval for H-mode is as fast as P-mode and the space efficiency is as good as C-mode.

• Proceedings of the KSPS conference
• /
• 2003.10a
• /
• pp.33-36
• /
• 2003

Two experiments were conducted to support the suggestion that the same information processing was used in both input modalities, visual and auditory modality in Wonil Choi & Kichun Nam(2003)'s paper. The primed lexical decision task was performed and pseudoword prime stimuli were used. The result was that priming effect did not occur in any experimental condition. This result might be interpreted visual facilitative information and phonological inhibitory information cancelled each other.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2003.06a
• /
• pp.46-49
• /
• 2003

The use of confidence measures for word/utterance verification has become art essential component of any speech input application. Confidence measures have applications to a number of problems such as rejection of incorrect hypotheses, speaker adaptation, or adaptive modification of the hypothesis score during search in continuous speech recognition. In this paper, we present a new utterance verification method using vowel string. Using subword HMMs of VCCV unit, we create anti-models which include vowel string in hypothesis words. The experiment results show that the utterance verification rate of the proposed method is about 79.5%.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.1705-1709
• /
• 2005

This paper covers review and challenges in the area of speech recognition by taking into account different classes of recognition mode. The recognition mode can be either speaker independent or speaker dependant. Size of the vocabulary and the input mode are two crucial factors for a speech recognizer. The input mode refers to continuous or isolated speech recognition system and the vocabulary size can be small less than hundred words or large less than few thousands words. This varies according to system design and objectives.[2]. The organization of the paper is: first it covers various fundamental methods of speech recognition, then it takes into account various deficiencies in the existing systems and finally it discloses the various probable application areas.

• Phonetics and Speech Sciences
• /
• v.2 no.4
• /
• pp.137-143
• /
• 2010

In this paper, a speech endpoint detection algorithm is proposed. The proposed algorithm is a kind of state transition-based ones for speech detection. To reject short-duration acoustic pulses which can be considered noises, it utilizes duration information of all detected pulses. For the optimization of parameters related with pulse lengths and energy threshold to detect speech intervals, an exhaustive search scheme is adopted while speech recognition rates are used as its performance index. Experimental results show that the proposed algorithm outperforms the baseline state-based endpoint detection algorithm. At 5 dB input SNR for the beamforming input, the word recognition accuracies of its outputs were 78.5% for human voice noises and 81.1% for music noises.