• Journal of Intelligence and Information Systems
• /
• v.21 no.1
• /
• pp.119-142
• /
• 2015

With the explosive growth in the volume of information, Internet users are experiencing considerable difficulties in obtaining necessary information online. Against this backdrop, ever-greater importance is being placed on a recommender system that provides information catered to user preferences and tastes in an attempt to address issues associated with information overload. To this end, a number of techniques have been proposed, including content-based filtering (CBF), demographic filtering (DF) and collaborative filtering (CF). Among them, CBF and DF require external information and thus cannot be applied to a variety of domains. CF, on the other hand, is widely used since it is relatively free from the domain constraint. The CF technique is broadly classified into memory-based CF, model-based CF and hybrid CF. Model-based CF addresses the drawbacks of CF by considering the Bayesian model, clustering model or dependency network model. This filtering technique not only improves the sparsity and scalability issues but also boosts predictive performance. However, it involves expensive model-building and results in a tradeoff between performance and scalability. Such tradeoff is attributed to reduced coverage, which is a type of sparsity issues. In addition, expensive model-building may lead to performance instability since changes in the domain environment cannot be immediately incorporated into the model due to high costs involved. Cumulative changes in the domain environment that have failed to be reflected eventually undermine system performance. This study incorporates the Markov model of transition probabilities and the concept of fuzzy clustering with CBCF to propose predictive clustering-based CF (PCCF) that solves the issues of reduced coverage and of unstable performance. The method improves performance instability by tracking the changes in user preferences and bridging the gap between the static model and dynamic users. Furthermore, the issue of reduced coverage also improves by expanding the coverage based on transition probabilities and clustering probabilities. The proposed method consists of four processes. First, user preferences are normalized in preference clustering. Second, changes in user preferences are detected from review score entries during preference transition detection. Third, user propensities are normalized using patterns of changes (propensities) in user preferences in propensity clustering. Lastly, the preference prediction model is developed to predict user preferences for items during preference prediction. The proposed method has been validated by testing the robustness of performance instability and scalability-performance tradeoff. The initial test compared and analyzed the performance of individual recommender systems each enabled by IBCF, CBCF, ICFEC and PCCF under an environment where data sparsity had been minimized. The following test adjusted the optimal number of clusters in CBCF, ICFEC and PCCF for a comparative analysis of subsequent changes in the system performance. The test results revealed that the suggested method produced insignificant improvement in performance in comparison with the existing techniques. In addition, it failed to achieve significant improvement in the standard deviation that indicates the degree of data fluctuation. Notwithstanding, it resulted in marked improvement over the existing techniques in terms of range that indicates the level of performance fluctuation. The level of performance fluctuation before and after the model generation improved by 51.31% in the initial test. Then in the following test, there has been 36.05% improvement in the level of performance fluctuation driven by the changes in the number of clusters. This signifies that the proposed method, despite the slight performance improvement, clearly offers better performance stability compared to the existing techniques. Further research on this study will be directed toward enhancing the recommendation performance that failed to demonstrate significant improvement over the existing techniques. The future research will consider the introduction of a high-dimensional parameter-free clustering algorithm or deep learning-based model in order to improve performance in recommendations.

• Journal of Intelligence and Information Systems
• /
• v.19 no.2
• /
• pp.39-54
• /
• 2013

Recent explosive increase of electronic commerce provides many advantageous purchase opportunities to customers. In this situation, customers who do not have enough knowledge about their purchases, may accept product recommendations. Product recommender systems automatically reflect user's preference and provide recommendation list to the users. Thus, product recommender system in online shopping store has been known as one of the most popular tools for one-to-one marketing. However, recommender systems which do not properly reflect user's preference cause user's disappointment and waste of time. In this study, we propose a novel recommender system which uses data mining and multi-model ensemble techniques to enhance the recommendation performance through reflecting the precise user's preference. The research data is collected from the real-world online shopping store, which deals products from famous art galleries and museums in Korea. The data initially contain 5759 transaction data, but finally remain 3167 transaction data after deletion of null data. In this study, we transform the categorical variables into dummy variables and exclude outlier data. The proposed model consists of two steps. The first step predicts customers who have high likelihood to purchase products in the online shopping store. In this step, we first use logistic regression, decision trees, and artificial neural networks to predict customers who have high likelihood to purchase products in each product group. We perform above data mining techniques using SAS E-Miner software. In this study, we partition datasets into two sets as modeling and validation sets for the logistic regression and decision trees. We also partition datasets into three sets as training, test, and validation sets for the artificial neural network model. The validation dataset is equal for the all experiments. Then we composite the results of each predictor using the multi-model ensemble techniques such as bagging and bumping. Bagging is the abbreviation of "Bootstrap Aggregation" and it composite outputs from several machine learning techniques for raising the performance and stability of prediction or classification. This technique is special form of the averaging method. Bumping is the abbreviation of "Bootstrap Umbrella of Model Parameter," and it only considers the model which has the lowest error value. The results show that bumping outperforms bagging and the other predictors except for "Poster" product group. For the "Poster" product group, artificial neural network model performs better than the other models. In the second step, we use the market basket analysis to extract association rules for co-purchased products. We can extract thirty one association rules according to values of Lift, Support, and Confidence measure. We set the minimum transaction frequency to support associations as 5%, maximum number of items in an association as 4, and minimum confidence for rule generation as 10%. This study also excludes the extracted association rules below 1 of lift value. We finally get fifteen association rules by excluding duplicate rules. Among the fifteen association rules, eleven rules contain association between products in "Office Supplies" product group, one rules include the association between "Office Supplies" and "Fashion" product groups, and other three rules contain association between "Office Supplies" and "Home Decoration" product groups. Finally, the proposed product recommender systems provides list of recommendations to the proper customers. We test the usability of the proposed system by using prototype and real-world transaction and profile data. For this end, we construct the prototype system by using the ASP, Java Script and Microsoft Access. In addition, we survey about user satisfaction for the recommended product list from the proposed system and the randomly selected product lists. The participants for the survey are 173 persons who use MSN Messenger, Daum Caf$\acute{e}$, and P2P services. We evaluate the user satisfaction using five-scale Likert measure. This study also performs "Paired Sample T-test" for the results of the survey. The results show that the proposed model outperforms the random selection model with 1% statistical significance level. It means that the users satisfied the recommended product list significantly. The results also show that the proposed system may be useful in real-world online shopping store.

• Korean Journal of Poultry Science
• /
• v.29 no.2
• /
• pp.95-100
• /
• 2002

Two experiments were conducted to investigate the effect of taurine supplementation on lipid metabolism in laying hens. In experiment 1, 19-wk-old laying hens were given one of four taurine supplemented diets (0 (control), 0.4, 0.8, and 1.2% taurine) fur 10 weeks. Abdominal fat weight was lower in the 1.2% diet by 29.2% compared to the control. Serum concentrations of triacylglycerol and HDL-cholesterol were not different among the treatments. However, seam concentration of total cholesterol was higher by 22.4% in the 1.2% diet compared to the control. Concentration of triacylglycerol or total cholesterol in the liver were decreased by 26.1% or 26.4% and 28.2% or 26.4%, respectively in the 0.8% and 1.2% diets compared to the control. The concentration of HDL-cholesterol in liver was also lower by 33.9% in the 1.2% diet compared to the control. In experiment 2, 81-wk-old laying hens were allocated to one of three taurine supplemented diets (0 (control), 1 and 2% taurine) fur 6 weeks. Abdominal fat weight was lower by 25% in 1% taurine supplementation compared to the control. Serum concentrations of triacylglycerol, total cholesterol and HDL-cholesterol of hens fed with 1% diet were not different from those of control group. However, sew concentrations of triacylglycerol and total cholesterol were lower by 44.0% and 19.8%, respectively in the 2% diet compared to the control. Furthermore, serum concentration of HDL -cholesterol in the 2% diet was higher by 75% compared to the control. Concentrations of triacylglycerol and total cholesterol in the liver in the 2% diet were decreased in the 1% diet by 36.8 and 23%, respectively, but increased by 78.4% and 70%, respectively, compared to the control. The concentration of HDL-cholesterol in the liver was not different between the 1% diet and the control, but higher by 62.8% in the 2% diet compared to the control. These results indicated that taurine supplementation decreased the fat storage in abdominal cavity, which was accompanied by the changes in triacylglycerol and cholesterol metabolisms of laying hens.

• Korean Journal of Soil Science and Fertilizer
• /
• v.33 no.2
• /
• pp.92-99
• /
• 2000

To compare the effect of food waste compost(FWC) application on the sodicity of paddy and upland soil, laboratory experiment was conducted. Six kinds of FWC made of various mixing ratio of food waste and pig slurry as raw material were applied to paddy soil under submerged condition and to upland soil in field water capacity, and were kept at $25^{\circ}C$ under laboratory incubation. The higher the mixing ratio of food waste on making FWC, the higher the FWC showed Na content and EC. Mineralized ratio of cations in FWC during incubation showed no difference between paddy and upland soil. It was high in the order of Na>K>Mg>Ca as 99, 94, 71, and 71%, respectively. NaCl contents of FWC applied to soils against SAR and ESP were fitted well to first linear regression with extremely high significance($R^2=0.99$). Increasing rate of SAR and ESP was higher in upland soil than paddy soil by 2.3 times. The difference was considered to be caused by dilution effect which was exerted by the application of more soil to water ratio to paddy soil than to upland soil on SAR analysis in consideration of cultivating condition. The calculated values of $([Ca^{2+}+Mg^{2+}]/2)^{1/2}$ used as a denominator on SAR calculation showed a little difference among FWC treatments by 2.1~2.4, while [$Na^+$] used as a numerator showed much variance by 3.1~9.5. Therefore, as a parameter for the assessment of FWC quality affecting soil sodicity, the use of only Na content in FWC was proposed without regarding Ca and Mg contents. Soil Ex. Na contents showed extremely high correlation($R^2=0.99$) with ESP. Moreover, because the former can be more easily determined than the latter, soil Ex. Na content was proposed as a new sodicity index.

• Journal of Food Hygiene and Safety
• /
• v.29 no.2
• /
• pp.131-140
• /
• 2014

The objective of this study was to develop a simultaneous method of 8 penicillin antibiotics including amoxicillin, ampicillin, cloxacillin, dicloxacillin, nafcillin, oxacillin, penicillin G and penicillin V in meat using LC-MS/MS. The procedure involves solid phase extraction with HLB cartridge and subsequent analysis by LC-MS/MS. To optimize MS analytical condition of 8 compounds, each parameter was established by multiple reaction monitoring in positive ion mode. The chromatographic separation was achieved on a $C_{18}$ column with a mobile phase of 0.05% formic acid and 0.05% formic acid in acetonitrile at a flow rate of 0.2 mL/min for 20 min with a gradient elution. The developed method was validated for specificity, linearity, accuracy and precision in beef, pork and chicken. The recoveries were 71.0~106%, and relative standard deviations (RSD) were 4.0~11.2%. The limit of detection (LOD) and the limit of quantification (LOQ) were 0.003~0.008 mg/kg and 0.01~0.03 mg/kg, respectively, that are below maximum residue limit (MRL) of the penicillins. This study also performed survey of residual penicillin antibiotics for 193 samples of beef, pork and chicken collected from 9 cities in Korea. Penicillins were not found in all the samples except a sample of pork which contained cloxacillin (concentration of 0.08 mg/kg) below the MRL (0.3 mg/kg).

• The Journal of Korean Society for Radiation Therapy
• /
• v.30 no.1_2
• /
• pp.177-183
• /
• 2018

Purpose : Volumetric Modulated Arc Therapy(VMAT) has the advantage of uniformly and precisely irradiating the tumor to the shape of the tumor while reducing the risk of radiation damage to normal tissues. such as brain cancer, head and neck cancer and prostate cancer, It is being used for treatment. The purpose of this study is to evaluate the usefulness of the Jaw-Tracking technique(JTT) in VMAT for brain and head and neck cancer. Materials and Methods : We selected eight patients with brain and head and neck cancer(4 Brain, 4 head and neck) who were treated with the VMAT treatment technique. Contouring information of the patient's tumor and normal organ was fused to the Rando phantom using the deformable registration of Velocity(Varian, USA). A treatment plan was developed using the Varian Eclipse(ver 15.5, Varian, USA) with the same patient actual beam parameters except for the use of jaw-tracking. As the evaluation index, the maximum dose and mean dose of target and OAR were compared and a portal dosimetry was performed for the treatment plan verification. Results : When using JTT, the relative dose of OAR decreased by 5.24 % and the maximum dose by 7.05 %, respectively, compared with the Static-Jaw technique(SJT). In the various OARs, the mean dose and maximum dose reduction ranges ranged from 0.01 to 3.16 Gy and from 0.12 to 6.27 Gy, respectively. In the case of the target, the maximum dose of GTV, CTV, PTV decreased by 0.17 %, 0.43 %, and 0.37 % in JTT, and the mean dose decreased by 0.24 %, 0.47 % and 0.47 %, respectively. Gamma analysis The JTT and SJT passing rates were $98{\pm}1.73%$ and $97{\pm}1.83%$ on the basis of 3 % / 3 mm, respectively. Comparing the doses of all OARs applied to the experiment, it was found that the use of JTT resulted in a significant decrease in dose due to additional jaw shielding besides MLC than SJT. Conclusion : In radiation therapy using VMAT treatment plan, we can apply JTT in the case of adjacent tumor and normal organs such as brain cancer and head and neck cancer, and in radiotherapy required large field and high energy caused increase leakage dose through MLC. It is considered that the target dose of PTV can be increased by lowering the dose of normal tissue surrounding the tumor.

• Korean journal of applied entomology
• /
• v.57 no.4
• /
• pp.261-269
• /
• 2018

The cowpea aphid Aphis craccivora Koch (Hemiptera: Aphididae) is a polyphagous species with a worldwide distribution. We investigated the temperature effects on development periods of nymphs, and the longevity and fecundity of apterous female of A. craccivora. The study was conducted at six constant temperatures of 10.0, 15.0, 20.0, 25, 30.0, and $32.5^{\circ}C$. A. craccivora developed successfully from nymph to adult stage at all temperatures subjected. The developmental rate of A. craccivora increased as temperature increased. The lower developmental threshold (LT) and thermal constant (K) of A. craccivora nymph stage were estimated by linear regression as $5.3^{\circ}C$ and 128.4 degree-days (DD), respectively. Lower and higher threshold temperatures (TL, TH and TH-TL, respectively) were calculated by the Sharpe_Schoolfield_Ikemoto (SSI) model as $17.0^{\circ}C$, $34.6^{\circ}C$ and $17.5^{\circ}C$. Developmental completion of nymph stages was described using a three-parameter Weibull function. Life table parameters were estimated. The intrinsic rate of increase was highest at $25^{\circ}C$, while the net reproductive rate was highest at $20^{\circ}C$. Biological characteristics of A. craccivora populations from different geographic areas were discussed.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.31 no.3
• /
• pp.158-169
• /
• 2019

A probabilistic model that can predict the residual useful lifetime of structure is formulated by using the gamma process which is one of the stochastic processes. The formulated stochastic model can take into account both the sampling uncertainty associated with damages measured up to now and the temporal uncertainty of cumulative damage over time. A method estimating several parameters of stochastic model is additionally proposed by introducing of the least square method and the method of moments, so that the age of a structure, the operational environment, and the evolution of damage with time can be considered. Some features related to the residual useful lifetime are firstly investigated into through the sensitivity analysis on parameters under a simple setting of single damage data measured at the current age. The stochastic model are then applied to the rubble-mound breakwater straightforwardly. The parameters of gamma process can be estimated for several experimental data on the damage processes of armor rocks of rubble-mound breakwater. The expected damage levels over time, which are numerically simulated with the estimated parameters, are in very good agreement with those from the flume testing. It has been found from various numerical calculations that the probabilities exceeding the failure limit are converged to the constraint that the model must be satisfied after lasting for a long time from now. Meanwhile, the expected residual useful lifetimes evaluated from the failure probabilities are seen to be different with respect to the behavior of damage history. As the coefficient of variation of cumulative damage is becoming large, in particular, it has been shown that the expected residual useful lifetimes have significant discrepancies from those of the deterministic regression model. This is mainly due to the effect of sampling and temporal uncertainties associated with damage, by which the first time to failure tends to be widely distributed. Therefore, the stochastic model presented in this paper for predicting the residual useful lifetime of structure can properly implement the probabilistic assessment on current damage state of structure as well as take account of the temporal uncertainty of future cumulative damage.

• Korean Journal of Mineralogy and Petrology
• /
• v.34 no.3
• /
• pp.193-208
• /
• 2021

The characteristics of the microcrack lengths(①), microcrack spacings(②) and Brazilian tensile strengths(③) related to the six directions of rock cleavages(H2~R1) in Geochang granite were analyzed. First, the 18 cumulative graphs for the above three major factors representing unique characteristics of the rock cleavages were made. Through the general chart for these graphs classified into three planes and three rock cleavages, the 28 parameters on the length, spacing and Brazilian tensile strength have been determined. The results of correlation analysis among these parameters are summarized as follows. Second, the above parameters were classified into six groups(I~VI) according to the sorting order on the magnitude of parameter values among three rock cleavages and three planes. The values of parameters belonging to group I and II are in order of R(rift) < G(grain) < H(hardway) and H < G < R. The values of the 8 parameters on the length of line(os₂, 𝚫s, 𝚫L and oS^mean), the exponent(λ_L^mean and λ_S^mean), the slope(a^mean) and the anisotropy coefficient (An^mean) are in order of R < G < H and H'(hardway plane) < G'(grain plane) < R'(rift plane). Third, the noticeable differences in distribution patterns among the six types of charts for three planes and three rock cleavages are as follows. From the chart for three planes, the values of 𝚫L, 𝚫s and 𝚫σ_t, corresponding to the distance between two points where the two fitting lines meet on the X-axis, increase in the order of R' < H' < G'. In particular, the two graphs of R2 and G2 related to the length and Brazilian tensile strength are almost parallel to each other and show the distribution characteristics of hardway plane. Among the graphs related to the Brazilian tensile strength, the overall shape for hardway plane is similar to that for grain. From the chart for three rock cleavages, the slopes of the graphs related to the length increase in the order of R < G < H, while those of the graphs related to the spacing and Brazilian tensile strength decrease in the order of R < G < H. Lastly, the characteristics of variation among the six rock cleavages, the three planes and the three rock cleavages were visualized through the correlation chart among the above parameters from this study.

• The Journal of Korean Society for Radiation Therapy
• /
• v.33
• /
• pp.89-97
• /
• 2021

Purpose: The purpose of this study is to compare and evaluate the dose change according to the gas volume variations in the rectum, which was not included in the treatment plan during radiation therapy for cervical cancer. Materials and methods: Static Intensity Modulated Radiation Therapy (S-IMRT) using a 9-field and Volumetric Modulated Arc Therapy (VMAT) using 2 full-arcs were established with treatment planning system on Computed Tomography images of a human phantom. Random gas parameters were included in the Planning Target Volume(PTV) with a maximum change of 2.0 cm in increments of 0.5 cm. Then, the Conformity Index (CI), Homogeneity Index (HI) and PTV D_max for the target volume were calculated, and the minimum dose (D_min), mean dose (D_mean) and Maximum Dose (D_max) were calculated and compared for OAR(organs at risk). For statistical analysis, T-test was performed to obtain a p-value, where the significance level was set to 0.05. Result: The HI coefficients of determination(R²) of S-IMRT and VMAT were 0.9423 and 0.8223, respectively, indicating a relatively clear correlation, and PTV D_max was found to increase up to 2.8% as the volume of a given gas parameter increased. In case of OAR evaluation, the dose in the bladder did not change with gas volume while a significant dose difference of more than D_mean 700 cGy was confirmed in rectum using both treatment plans at gas volumes of 1.0 cm or more. In all values except for D_mean of bladder, p-value was less than 0.05, confirming a statistically significant difference. Conclusion: In the case of gas generation not considered in the reference treatment plan, as the amount of gas increased, the dose difference at PTV and the dose delivered to the rectum increased. Therefore, during radiation therapy, it is necessary to make efforts to minimize the dose transmission error caused by a large amount of gas volumes in the rectum. Further studies will be necessary to evaluate dose transmission by not only varying the gas volume but also where the gas was located in the treatment field.