• The Journal of the Acoustical Society of Korea
• /
• v.21 no.3
• /
• pp.274-284
• /
• 2002

Word segmentation errors occurring in text preprocessing often insert incorrect words into recognition vocabulary and cause poor language models for Korean large vocabulary continuous speech recognition. We propose an automatic word segmentation algorithm using Markov chains and syllable-based n-gram language models in order to correct word segmentation error in teat corpora. We assume that a sentence is generated from a Markov chain. Spaces and non-space characters are generated on self-transitions and other transitions of the Markov chain, respectively Then word segmentation of the sentence is obtained by finding the maximum likelihood path using syllable n-gram scores. In experimental results, the algorithm showed 91.58% word accuracy and 96.69% syllable accuracy for word segmentation of 254 sentence newspaper columns without any spaces. The algorithm improved the word accuracy from 91.00% to 96.27% for word segmentation correction at line breaks and yielded the decomposition accuracy of 96.22% for compound-noun decomposition.

• The KIPS Transactions:PartB
• /
• v.13B no.1 s.104
• /
• pp.71-76
• /
• 2006

To obtain a correctly translated sentence in a machine translation system, we must select target words that not only reflect an appropriate meaning in a source sentence but also make a fluent sentence in a target language. This paper points out that a source language word has various senses and each sense can be mapped into multiple target words, and proposes a new translation disambiguation method based on this 'word-to-sense and sense-to-word' relationship. In my method target words are chosen through disambiguation of a source word sense and selection of a target word. Most of translation disambiguation methods are based on a 'word-to-word' relationship that means they translate a source word directly into a target wort so they require complicate knowledge sources that directly link a source words to target words, which are hard to obtain like bilingual aligned corpora. By combining two sub-problems for each language, knowledge for translation disambiguation can be automatically extracted from knowledge sources for each language that are easy to obtain. In addition, disambiguation results satisfy both fidelity and intelligibility because selected target words have correct meaning and generate naturally composed target sentences.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.8
• /
• pp.3168-3186
• /
• 2020

Automatic topic essay generation (TEG) is a controllable text generation task that aims to generate informative, diverse, and topic-consistent essays based on multiple topics. To make the generated essays of high quality, a reasonable method should consider both diversity and topic-consistency. Another essential issue is the intrinsic link of the topics, which contributes to making the essays closely surround the semantics of provided topics. However, it remains challenging for TEG to fill the semantic gap between source topic words and target output, and a more powerful model is needed to capture the semantics of given topics. To this end, we propose a pretraining-based contextual self-attention (PC-SAN) model that is built upon the seq2seq framework. For the encoder of our model, we employ a dynamic weight sum of layers from BERT to fully utilize the semantics of topics, which is of great help to fill the gap and improve the quality of the generated essays. In the decoding phase, we also transform the target-side contextual history information into the query layers to alleviate the lack of context in typical self-attention networks (SANs). Experimental results on large-scale paragraph-level Chinese corpora verify that our model is capable of generating diverse, topic-consistent text and essentially makes improvements as compare to strong baselines. Furthermore, extensive analysis validates the effectiveness of contextual embeddings from BERT and contextual history information in SANs.

• Asian-Australasian Journal of Animal Sciences
• /
• v.13 no.5
• /
• pp.595-599
• /
• 2000

Fine wool sheep (n=18) maintained in a tropical environment were allocated to three treatment groups. Estrus was induced with two injections of $PGF_{2{\alpha}}$ (10 mg. im) at 10 days interval. Superovulation treatment started 2 days prior to the second injection of $PGF_{2{\alpha}}$. Each ewe was treated with a total dose of 25 units FSH (Super-OV) i.m. every 12 hover 3 days; Group 2 were also injected i.m. with 200 IU PMSG at the first injection of FSH; Group 3 was treated as in Group 2 and also with GnRH ($4{\mu}g$ Buserelin) at the onset of estrus. The ewes in estrus were mated with a fertile ram. Ovarian examination and recovery of embryo and ova were performed at laparoscopy and laparotomy on day 3 or 4 after mating. Data for onset of estrus, duration of estrus, number of corpora lutea (CL), number of unnovulated large follicle (LF), embryo recovery rate, embryo quality and fertilization recorded for the 3 groups. Ewes in the Group 1 set in estrus later (p<0.05; $50.0{\pm}7.29h$) than the ewes in Group 2 ($24.5{\pm}3.58$) and 3 ($32.5{\pm}3.58h$). The duration of estrus, ovarian size and ovarian response (number of CL and LF) did not differ significantly (p>0.05) among the 3 groups. The proportion of ewes with a superovulatory response (${\geq}2$ CL) was the lowest (50%) in Group 1 treated with FSH alone but ova/embryo recovery (100%) and fertilization (100%) was significantly (p<0.05) higher than Group 2 (58.3 and 85.7%, respectively) and Group 3 (48.6 and 50%, respectively). It is concluded that in tropical fine wool sheep, there is no difference in the 3 treatments for yield of good quality embryos but ovarian response and ovulation rate increased on additional use of PMSG and GnRH respectively to FSH alone.

• Korean journal of applied entomology
• /
• v.35 no.3
• /
• pp.238-242
• /
• 1996

The second instar milkweed bugs, Oncopeltus fasciatus were exposed to the residue of 6-methoxy-7-ethoxy-2, 2- dimethylchromene(or 7-ethoxy precocene 11, 7-EP-11) with one, two, or four different kinds of synergist(s) deposited as a residue on 9 cm diameter petri dishes. The anti juvenile hormone analog, 7-EP-I1 yielded a 50% effectiveness concentration at 1.18 wg/cmZ. The concentration of 7-EP-I1 and R020-9747 combined gave 0.084 ~*g/cm', as much 14 times more potent than with 7-EP-I1 alone. Most of the other synergists, including geraniol, 4-chalcone oxide, isosafrole, and piperonyl butoxide however, showed comparatively low level synergism with 1.0-4.3 times, depending upon the combinations of the synergist(s). The obtained results were considered that R020-9747, an analog of benzopyran with similar structure to that of 7-EP-I1 inhibited selectively the insect monooxygenase oxidative defense mechanism, in the body except for corpora allata because general antioxidants of piperonyl butoxide, isosafrole, and 4-phenyl chalcone oxide showed relatively mild synergism.

• The Journal of Korean Obstetrics and Gynecology
• /
• v.23 no.2
• /
• pp.1-19
• /
• 2010

Purpose: This study was designed to investigate the effects of Gaeullijin-Tang(GIT) on the progression of the estradiol valerate(EV)-induced polycystic ovaries(PCO) in rats. Methods: PCO was induced by single intramuscular injection with estradiol valerate(EV)(4mg) in female rats. Normal group(n=8) were injected with sesame oil and orally administrated distilled water for sixty days. PCO control group (n=8) were injected with EV and orally administrated distilled water for sixty days. GIT treated group(n=8) were injected with EV and orally administrated GIT for same duration. At the end day of experiment, we measured weights of body, ovaries, adrenal glands, uterus and contents of serum androsternedione (ADD) and total estrogen. The histomorphometrical changes of ovaries were also evaluated. And we observed the NGF and CRF expression by immunohistochemistry. Results: The results were as follows - The weights(mg) of ovaries in GIT treated group($75.5{\pm}4.3$) were significantly increased(p<0.001) compared with PCO control group($54.3{\pm}4.5$). - The numbers of mature follicles in GIT treated group($5.3{\pm}1.2$) were significantly increased(p<0.05) compared with PCO control group($3.5{\pm}1.2$). - The numbers of corpora lutea in GIT treated group($1.8{\pm}1.3$) were significantly increased(p<0.05) compared with PCO control group($0.3{\pm}0.7$). Conclusion: From the above results, we concluded that Gaeullijin-Tang(GIT) contributes to stimulating normally ovulation by decreasesing ovarian NGF concentrations.

• The Journal of Korean Obstetrics and Gynecology
• /
• v.21 no.2
• /
• pp.1-16
• /
• 2008

Purpose: In this rearch, we made a plan to investigate the effects of Changbuyeukgun-Tang(CET) on the polycystic ovary(PCOS) induced by estradiol valerate(EV) in rats. PCO was induced by single intramuscular injection with EV(4㎎) in female rats. Rats(n=8) in normal group were injected with sesame oil and orally administrated distilled water for 8weeks Rats in PCO control group(n=8) were injected with EV and orally administrated distilled water for 8weeks. Rats in CET treated group(n=8) were injected with EV and orally administrated CET for 8weeks. Methods: We measured the weights of the body, the ovary, the uterus and the adrenal gland. And also, we analyzed ovarian histopathology, NGF and CRF immunohistochemistry. Results: The results were as follows1. the weights(㎎) of the ovaries in CET treated group were significantly increased(p<0.001) compared with PCO control group. 2. the number of the mature follicles in CET treated group were significantly increased(p<0.01) compared with PCO control. 3. the number of the atresia follicles in CET treated group were significantly decreased(p<0.01) compared with PCO control. 4. the number of the corpora in CET treated group were significantly increased(p<0.05) compared with PCO control. 5. the expressions of NGF-immunoreactive cells in the ovarian granulosa cells of rats in CET treated group were less observed than PCO control group. Conclusion: According to these results, we finally concluded that Changbuyeukgun-Tang(CET) has the inhibitory effect on the development of EV-induced polycystic ovary. And we deduced that the effect of it may be related to the decreased NGF activities in the ovary.

• Development and Reproduction
• /
• v.20 no.2
• /
• pp.91-99
• /
• 2016

Lipopolysaccharide (LPS), an endotoxin, elicits strong immune responses in mammals. Several lines of evidence demonstrate that LPS challenge profoundly affects female reproductive function. For example, LPS exposure affects steroidogenesis and folliculogenesis, resulting in delayed puberty onset. The present study was conducted to clarify the mechanism underlying the adverse effect of LPS on the delayed puberty in female rats. LPS was daily injected for 5 days ($50{\mu}g/kg$, PND 25-29) to treated animals and the date at VO was evaluated through daily visual examination. At PND 39, animals were sacrificed, and the tissues were immediately removed and weighed. Among the reproductive organs, the weights of the ovaries and oviduct from LPS-treated animals were significantly lower than those of control animals. There were no changes in the weights of uterus and vagina between the LPS-treated and their control animals. immunological challenge by LPS delayed VO. Multiple corpora lutea were found in the control ovaries, indicating ovulations were occurred. However, none of corpus luteum was present in the LPS-treated ovary. The transcription level of steroidogenic acute regulatory protein (StAR), CYP11A1, CYP17A1 and CYP19 were significantly increased by LPS treatment. On the other hand, the levels of $3{\beta}$-HSD, $17{\beta}$-HSD and LH receptor were not changed by LPS challenge. In conclusion, the present study demonstrated that the repeated LPS exposure during the prepubertal period could induce multiple alterations in the steroidogenic machinery in ovary, and in turn, delayed puberty onset. The prepubertal LPS challenge model used in our study is useful to understand the reciprocal regulation of immune (stress) - reproductive function in early life.

• Asian-Australasian Journal of Animal Sciences
• /
• v.16 no.3
• /
• pp.342-347
• /
• 2003

The enzyme $20{\alpha}$-hydroxysteroid dehydrogenase ($20{\alpha}$-HSD) catabolizes progesterone to $20{\alpha}$-dihydroprogesterone ($20{\alpha}$-OHP), and is appeared in rat corpora luteal and placenta. A polled samples of 10-15 placental and ovarian tissues collected from each individual rat were subjected to measurement of $20{\alpha}$-HSD activity. A $20{\alpha}$-HSD activity in the cytosol fraction was based on the generation of NADPH. In this study, it is designed to study cytosolic $20{\alpha}$-HSD activity in rat ovarian and placenta during pregnancy, and its relationship to embryonic mortality. It was found that from days 5 to 18 of pregnancy the $20{\alpha}$-HSD activities steady by decreased but at parturition time rapidly increased in ovary. On the other hand, placental cytosolic $20{\alpha}$-HSD activities were high detected from days 8 to 10 of pregnancy, not detectable from days 11 to 20 of pregnancy, but again very high at the time of parturition. Analysis of DEAE column chromatography revealed that two different types of $20{\alpha}$-HSD (HSD-1 and HSD-2) were found with similar activity in the placental cytosol on day 10 of pregnancy. The number of fetuses on day 10 of pregnancy was 15.4 and decreased significantly to 12.9 on day 12. The results suggested that expression of $20{\alpha}$-HSD in the placental tissues seems to be related the number of fetal survived in the specific time (days 11 and 12) which spontaneous fetal loss occurs.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.41 no.5
• /
• pp.592-601
• /
• 2016

The recurrent neural network (RNN) is a deep learning model which is suitable to sequential or length-variable data. The Long Short-Term Memory (LSTM) mitigates the vanishing gradient problem of RNNs so that LSTM can maintain the long-term dependency among the constituents of the given input sequence. In this paper, we propose a LSTM based language model which can predict following words of a given incomplete sentence to generate a complete sentence. To evaluate our method, we trained our model using multiple Korean corpora then generated the incomplete part of Korean sentences. The result shows that our language model was able to generate the fluent Korean sentences. We also show that the word based model generated better sentences compared to the other settings.