• Journal of Intelligence and Information Systems
• /
• v.27 no.1
• /
• pp.83-102
• /
• 2021

The government recently announced various policies for developing big-data and artificial intelligence fields to provide a great opportunity to the public with respect to disclosure of high-quality data within public institutions. KSURE(Korea Trade Insurance Corporation) is a major public institution for financial policy in Korea, and thus the company is strongly committed to backing export companies with various systems. Nevertheless, there are still fewer cases of realized business model based on big-data analyses. In this situation, this paper aims to develop a new business model which can be applied to an ex-ante prediction for the likelihood of the insurance accident of credit guarantee. We utilize internal data from KSURE which supports export companies in Korea and apply machine learning models. Then, we conduct performance comparison among the predictive models including Logistic Regression, Random Forest, XGBoost, LightGBM, and DNN(Deep Neural Network). For decades, many researchers have tried to find better models which can help to predict bankruptcy since the ex-ante prediction is crucial for corporate managers, investors, creditors, and other stakeholders. The development of the prediction for financial distress or bankruptcy was originated from Smith(1930), Fitzpatrick(1932), or Merwin(1942). One of the most famous models is the Altman's Z-score model(Altman, 1968) which was based on the multiple discriminant analysis. This model is widely used in both research and practice by this time. The author suggests the score model that utilizes five key financial ratios to predict the probability of bankruptcy in the next two years. Ohlson(1980) introduces logit model to complement some limitations of previous models. Furthermore, Elmer and Borowski(1988) develop and examine a rule-based, automated system which conducts the financial analysis of savings and loans. Since the 1980s, researchers in Korea have started to examine analyses on the prediction of financial distress or bankruptcy. Kim(1987) analyzes financial ratios and develops the prediction model. Also, Han et al.(1995, 1996, 1997, 2003, 2005, 2006) construct the prediction model using various techniques including artificial neural network. Yang(1996) introduces multiple discriminant analysis and logit model. Besides, Kim and Kim(2001) utilize artificial neural network techniques for ex-ante prediction of insolvent enterprises. After that, many scholars have been trying to predict financial distress or bankruptcy more precisely based on diverse models such as Random Forest or SVM. One major distinction of our research from the previous research is that we focus on examining the predicted probability of default for each sample case, not only on investigating the classification accuracy of each model for the entire sample. Most predictive models in this paper show that the level of the accuracy of classification is about 70% based on the entire sample. To be specific, LightGBM model shows the highest accuracy of 71.1% and Logit model indicates the lowest accuracy of 69%. However, we confirm that there are open to multiple interpretations. In the context of the business, we have to put more emphasis on efforts to minimize type 2 error which causes more harmful operating losses for the guaranty company. Thus, we also compare the classification accuracy by splitting predicted probability of the default into ten equal intervals. When we examine the classification accuracy for each interval, Logit model has the highest accuracy of 100% for 0~10% of the predicted probability of the default, however, Logit model has a relatively lower accuracy of 61.5% for 90~100% of the predicted probability of the default. On the other hand, Random Forest, XGBoost, LightGBM, and DNN indicate more desirable results since they indicate a higher level of accuracy for both 0~10% and 90~100% of the predicted probability of the default but have a lower level of accuracy around 50% of the predicted probability of the default. When it comes to the distribution of samples for each predicted probability of the default, both LightGBM and XGBoost models have a relatively large number of samples for both 0~10% and 90~100% of the predicted probability of the default. Although Random Forest model has an advantage with regard to the perspective of classification accuracy with small number of cases, LightGBM or XGBoost could become a more desirable model since they classify large number of cases into the two extreme intervals of the predicted probability of the default, even allowing for their relatively low classification accuracy. Considering the importance of type 2 error and total prediction accuracy, XGBoost and DNN show superior performance. Next, Random Forest and LightGBM show good results, but logistic regression shows the worst performance. However, each predictive model has a comparative advantage in terms of various evaluation standards. For instance, Random Forest model shows almost 100% accuracy for samples which are expected to have a high level of the probability of default. Collectively, we can construct more comprehensive ensemble models which contain multiple classification machine learning models and conduct majority voting for maximizing its overall performance.

• Journal of Intelligence and Information Systems
• /
• v.22 no.1
• /
• pp.19-41
• /
• 2016

Recently, the proliferation of mobile devices such as smartphones and tablet personal computers and the development of information communication technologies (ICT) have led to a big trend of a shift from single-channel shopping to multi-channel shopping. With the emergence of a "smart" group of consumers who want to shop in more reasonable and convenient ways, the boundaries apparently dividing online and offline shopping have collapsed and blurred more than ever before. Thus, there is now fierce competition between online and offline channels. Ever since the emergence of online shopping, a major type of multi-channel shopping has been "showrooming," where consumers visit offline stores to examine products before buying them online. However, because of the growing use of smart devices and the counterattack of offline retailers represented by omni-channel marketing strategies, one of the latest huge trends of shopping is "webrooming," where consumers visit online stores to examine products before buying them offline. This has become a threat to online retailers. In this situation, although it is very important to examine the influencing factors for switching from online shopping to webrooming, most prior studies have mainly focused on a single- or multi-channel shopping pattern. Therefore, this study thoroughly investigated the influencing factors on customers switching from online shopping to webrooming in terms of both the "search" and "purchase" processes through the application of a push-pull-mooring (PPM) framework. In order to test the research model, 280 individual samples were gathered from undergraduate and graduate students who had actual experience with webrooming. The results of the structural equation model (SEM) test revealed that the "pull" effect is strongest on the webrooming intention rather than the "push" or "mooring" effects. This proves a significant relationship between "attractiveness of webrooming" and "webrooming intention." In addition, the results showed that both the "perceived risk of online search" and "perceived risk of online purchase" significantly affect "distrust of online shopping." Similarly, both "perceived benefit of multi-channel search" and "perceived benefit of offline purchase" were found to have significant effects on "attractiveness of webrooming" were also found. Furthermore, the results indicated that "online purchase habit" is the only influencing factor that leads to "online shopping lock-in." The theoretical implications of the study are as follows. First, by examining the multi-channel shopping phenomenon from the perspective of "shopping switching" from online shopping to webrooming, this study complements the limits of the "channel switching" perspective, represented by multi-channel freeriding studies that merely focused on customers' channel switching behaviors from one to another. While extant studies with a channel switching perspective have focused on only one type of multi-channel shopping, where consumers just move from one particular channel to different channels, a study with a shopping switching perspective has the advantage of comprehensively investigating how consumers choose and navigate among diverse types of single- or multi-channel shopping alternatives. In this study, only limited shopping switching behavior from online shopping to webrooming was examined; however, the results should explain various phenomena in a more comprehensive manner from the perspective of shopping switching. Second, this study extends the scope of application of the push-pull-mooring framework, which is quite commonly used in marketing research to explain consumers' product switching behaviors. Through the application of this framework, it is hoped that more diverse shopping switching behaviors can be examined in future research. This study can serve a stepping stone for future studies. One of the most important practical implications of the study is that it may help single- and multi-channel retailers develop more specific customer strategies by revealing the influencing factors of webrooming intention from online shopping. For example, online single-channel retailers can ease the distrust of online shopping to prevent consumers from churning by reducing the perceived risk in terms of online search and purchase. On the other hand, offline retailers can develop specific strategies to increase the attractiveness of webrooming by letting customers perceive the benefits of multi-channel search or offline purchase. Although this study focused only on customers switching from online shopping to webrooming, the results can be expanded to various types of shopping switching behaviors embedded in single- and multi-channel shopping environments, such as showrooming and mobile shopping.

• Journal of Intelligence and Information Systems
• /
• v.22 no.3
• /
• pp.45-69
• /
• 2016

Recently, sentiment analysis using open Internet data is actively performed for various purposes. As online Internet communication channels become popular, companies try to capture public sentiment of them from online open information sources. This research is conducted for the purpose of analyzing pulbic sentiment of Korean Top-10 companies using a multi-categorical sentiment lexicon. Whereas existing researches related to public sentiment measurement based on big data approach classify sentiment into dimensions, this research classifies public sentiment into multiple categories. Dimensional sentiment structure has been commonly applied in sentiment analysis of various applications, because it is academically proven, and has a clear advantage of capturing degree of sentiment and interrelation of each dimension. However, the dimensional structure is not effective when measuring public sentiment because human sentiment is too complex to be divided into few dimensions. In addition, special training is needed for ordinary people to express their feeling into dimensional structure. People do not divide their sentiment into dimensions, nor do they need psychological training when they feel. People would not express their feeling in the way of dimensional structure like positive/negative or active/passive; rather they express theirs in the way of categorical sentiment like sadness, rage, happiness and so on. That is, categorial approach of sentiment analysis is more natural than dimensional approach. Accordingly, this research suggests multi-categorical sentiment structure as an alternative way to measure social sentiment from the point of the public. Multi-categorical sentiment structure classifies sentiments following the way that ordinary people do although there are possibility to contain some subjectiveness. In this research, nine categories: 'Sadness', 'Anger', 'Happiness', 'Disgust', 'Surprise', 'Fear', 'Interest', 'Boredom' and 'Pain' are used as multi-categorical sentiment structure. To capture public sentiment of Korean Top-10 companies, Internet news data of the companies are collected over the past 25 months from a representative Korean portal site. Based on the sentiment words extracted from previous researches, we have created a sentiment lexicon, and analyzed the frequency of the words coming up within the news data. The frequency of each sentiment category was calculated as a ratio out of the total sentiment words to make ranks of distributions. Sentiment comparison among top-4 companies, which are 'Samsung', 'Hyundai', 'SK', and 'LG', were separately visualized. As a next step, the research tested hypothesis to prove the usefulness of the multi-categorical sentiment lexicon. It tested how effective categorial sentiment can be used as relative comparison index in cross sectional and time series analysis. To test the effectiveness of the sentiment lexicon as cross sectional comparison index, pair-wise t-test and Duncan test were conducted. Two pairs of companies, 'Samsung' and 'Hanjin', 'SK' and 'Hanjin' were chosen to compare whether each categorical sentiment is significantly different in pair-wise t-test. Since category 'Sadness' has the largest vocabularies, it is chosen to figure out whether the subgroups of the companies are significantly different in Duncan test. It is proved that five sentiment categories of Samsung and Hanjin and four sentiment categories of SK and Hanjin are different significantly. In category 'Sadness', it has been figured out that there were six subgroups that are significantly different. To test the effectiveness of the sentiment lexicon as time series comparison index, 'nut rage' incident of Hanjin is selected as an example case. Term frequency of sentiment words of the month when the incident happened and term frequency of the one month before the event are compared. Sentiment categories was redivided into positive/negative sentiment, and it is tried to figure out whether the event actually has some negative impact on public sentiment of the company. The difference in each category was visualized, moreover the variation of word list of sentiment 'Rage' was shown to be more concrete. As a result, there was huge before-and-after difference of sentiment that ordinary people feel to the company. Both hypotheses have turned out to be statistically significant, and therefore sentiment analysis in business area using multi-categorical sentiment lexicons has persuasive power. This research implies that categorical sentiment analysis can be used as an alternative method to supplement dimensional sentiment analysis when figuring out public sentiment in business environment.

• Journal of Intelligence and Information Systems
• /
• v.22 no.3
• /
• pp.143-163
• /
• 2016

The demographics of Internet users are the most basic and important sources for target marketing or personalized advertisements on the digital marketing channels which include email, mobile, and social media. However, it gradually has become difficult to collect the demographics of Internet users because their activities are anonymous in many cases. Although the marketing department is able to get the demographics using online or offline surveys, these approaches are very expensive, long processes, and likely to include false statements. Clickstream data is the recording an Internet user leaves behind while visiting websites. As the user clicks anywhere in the webpage, the activity is logged in semi-structured website log files. Such data allows us to see what pages users visited, how long they stayed there, how often they visited, when they usually visited, which site they prefer, what keywords they used to find the site, whether they purchased any, and so forth. For such a reason, some researchers tried to guess the demographics of Internet users by using their clickstream data. They derived various independent variables likely to be correlated to the demographics. The variables include search keyword, frequency and intensity for time, day and month, variety of websites visited, text information for web pages visited, etc. The demographic attributes to predict are also diverse according to the paper, and cover gender, age, job, location, income, education, marital status, presence of children. A variety of data mining methods, such as LSA, SVM, decision tree, neural network, logistic regression, and k-nearest neighbors, were used for prediction model building. However, this research has not yet identified which data mining method is appropriate to predict each demographic variable. Moreover, it is required to review independent variables studied so far and combine them as needed, and evaluate them for building the best prediction model. The objective of this study is to choose clickstream attributes mostly likely to be correlated to the demographics from the results of previous research, and then to identify which data mining method is fitting to predict each demographic attribute. Among the demographic attributes, this paper focus on predicting gender, age, marital status, residence, and job. And from the results of previous research, 64 clickstream attributes are applied to predict the demographic attributes. The overall process of predictive model building is compose of 4 steps. In the first step, we create user profiles which include 64 clickstream attributes and 5 demographic attributes. The second step performs the dimension reduction of clickstream variables to solve the curse of dimensionality and overfitting problem. We utilize three approaches which are based on decision tree, PCA, and cluster analysis. We build alternative predictive models for each demographic variable in the third step. SVM, neural network, and logistic regression are used for modeling. The last step evaluates the alternative models in view of model accuracy and selects the best model. For the experiments, we used clickstream data which represents 5 demographics and 16,962,705 online activities for 5,000 Internet users. IBM SPSS Modeler 17.0 was used for our prediction process, and the 5-fold cross validation was conducted to enhance the reliability of our experiments. As the experimental results, we can verify that there are a specific data mining method well-suited for each demographic variable. For example, age prediction is best performed when using the decision tree based dimension reduction and neural network whereas the prediction of gender and marital status is the most accurate by applying SVM without dimension reduction. We conclude that the online behaviors of the Internet users, captured from the clickstream data analysis, could be well used to predict their demographics, thereby being utilized to the digital marketing.

• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.123-138
• /
• 2017

Since the stock market is driven by the expectation of traders, studies have been conducted to predict stock price movements through analysis of various sources of text data. In order to predict stock price movements, research has been conducted not only on the relationship between text data and fluctuations in stock prices, but also on the trading stocks based on news articles and social media responses. Studies that predict the movements of stock prices have also applied classification algorithms with constructing term-document matrix in the same way as other text mining approaches. Because the document contains a lot of words, it is better to select words that contribute more for building a term-document matrix. Based on the frequency of words, words that show too little frequency or importance are removed. It also selects words according to their contribution by measuring the degree to which a word contributes to correctly classifying a document. The basic idea of constructing a term-document matrix was to collect all the documents to be analyzed and to select and use the words that have an influence on the classification. In this study, we analyze the documents for each individual item and select the words that are irrelevant for all categories as neutral words. We extract the words around the selected neutral word and use it to generate the term-document matrix. The neutral word itself starts with the idea that the stock movement is less related to the existence of the neutral words, and that the surrounding words of the neutral word are more likely to affect the stock price movements. And apply it to the algorithm that classifies the stock price fluctuations with the generated term-document matrix. In this study, we firstly removed stop words and selected neutral words for each stock. And we used a method to exclude words that are included in news articles for other stocks among the selected words. Through the online news portal, we collected four months of news articles on the top 10 market cap stocks. We split the news articles into 3 month news data as training data and apply the remaining one month news articles to the model to predict the stock price movements of the next day. We used SVM, Boosting and Random Forest for building models and predicting the movements of stock prices. The stock market opened for four months (2016/02/01 ~ 2016/05/31) for a total of 80 days, using the initial 60 days as a training set and the remaining 20 days as a test set. The proposed word - based algorithm in this study showed better classification performance than the word selection method based on sparsity. This study predicted stock price volatility by collecting and analyzing news articles of the top 10 stocks in market cap. We used the term - document matrix based classification model to estimate the stock price fluctuations and compared the performance of the existing sparse - based word extraction method and the suggested method of removing words from the term - document matrix. The suggested method differs from the word extraction method in that it uses not only the news articles for the corresponding stock but also other news items to determine the words to extract. In other words, it removed not only the words that appeared in all the increase and decrease but also the words that appeared common in the news for other stocks. When the prediction accuracy was compared, the suggested method showed higher accuracy. The limitation of this study is that the stock price prediction was set up to classify the rise and fall, and the experiment was conducted only for the top ten stocks. The 10 stocks used in the experiment do not represent the entire stock market. In addition, it is difficult to show the investment performance because stock price fluctuation and profit rate may be different. Therefore, it is necessary to study the research using more stocks and the yield prediction through trading simulation.

• Journal of Intelligence and Information Systems
• /
• v.21 no.2
• /
• pp.191-204
• /
• 2015

According to the '2013 Performance Assessment Report on the Financial Program' from the National Assembly Budget Office, the unfilled recruitment ratio of Software(SW) Developers in South Korea was 25% in the 2012 fiscal year. Moreover, the unfilled recruitment ratio of highly-qualified SW developers reaches almost 80%. This phenomenon is intensified in small and medium enterprises consisting of less than 300 employees. Young job-seekers in South Korea are increasingly avoiding becoming a SW developer and even the current SW developers want to change careers, which hinders the national development of IT industries. The Korean government has recently realized the problem and implemented policies to foster young SW developers. Due to this effort, it has become easier to find young SW developers at the beginning-level. However, it is still hard to recruit highly-qualified SW developers for many IT companies. This is because in order to become a SW developing expert, having a long term experiences are important. Thus, improving job continuity intentions of current SW developers is more important than fostering new SW developers. Therefore, this study surveyed the job continuity intentions of SW developers and analyzed the factors associated with them. As a method, we carried out a survey from September 2014 to October 2014, which was targeted on 130 SW developers who were working in IT industries in South Korea. We gathered the demographic information and characteristics of the respondents, work environments of a SW industry, and social positions for SW developers. Afterward, a regression analysis and a decision tree method were performed to analyze the data. These two methods are widely used data mining techniques, which have explanation ability and are mutually complementary. We first performed a linear regression method to find the important factors assaociated with a job continuity intension of SW developers. The result showed that an 'expected age' to work as a SW developer were the most significant factor associated with the job continuity intention. We supposed that the major cause of this phenomenon is the structural problem of IT industries in South Korea, which requires SW developers to change the work field from developing area to management as they are promoted. Also, a 'motivation' to become a SW developer and a 'personality (introverted tendency)' of a SW developer are highly importantly factors associated with the job continuity intention. Next, the decision tree method was performed to extract the characteristics of highly motivated developers and the low motivated ones. We used well-known C4.5 algorithm for decision tree analysis. The results showed that 'motivation', 'personality', and 'expected age' were also important factors influencing the job continuity intentions, which was similar to the results of the regression analysis. In addition to that, the 'ability to learn' new technology was a crucial factor for the decision rules of job continuity. In other words, a person with high ability to learn new technology tends to work as a SW developer for a longer period of time. The decision rule also showed that a 'social position' of SW developers and a 'prospect' of SW industry were minor factors influencing job continuity intensions. On the other hand, 'type of an employment (regular position/ non-regular position)' and 'type of company (ordering company/ service providing company)' did not affect the job continuity intension in both methods. In this research, we demonstrated the job continuity intentions of SW developers, who were actually working at IT companies in South Korea, and we analyzed the factors associated with them. These results can be used for human resource management in many IT companies when recruiting or fostering highly-qualified SW experts. It can also help to build SW developer fostering policy and to solve the problem of unfilled recruitment of SW Developers in South Korea.

• Journal of Intelligence and Information Systems
• /
• v.18 no.4
• /
• pp.59-77
• /
• 2012

This study focuses on predicting which firms will increase capital by issuing new stocks in the near future. Many stakeholders, including banks, credit rating agencies and investors, performs a variety of analyses for firms' growth, profitability, stability, activity, productivity, etc., and regularly report the firms' financial analysis indices. In the paper, we develop predictive models for rights issues using these financial analysis indices and data mining techniques. This study approaches to building the predictive models from the perspective of two different analyses. The first is the analysis period. We divide the analysis period into before and after the IMF financial crisis, and examine whether there is the difference between the two periods. The second is the prediction time. In order to predict when firms increase capital by issuing new stocks, the prediction time is categorized as one year, two years and three years later. Therefore Total six prediction models are developed and analyzed. In this paper, we employ the decision tree technique to build the prediction models for rights issues. The decision tree is the most widely used prediction method which builds decision trees to label or categorize cases into a set of known classes. In contrast to neural networks, logistic regression and SVM, decision tree techniques are well suited for high-dimensional applications and have strong explanation capabilities. There are well-known decision tree induction algorithms such as CHAID, CART, QUEST, C5.0, etc. Among them, we use C5.0 algorithm which is the most recently developed algorithm and yields performance better than other algorithms. We obtained data for the rights issue and financial analysis from TS2000 of Korea Listed Companies Association. A record of financial analysis data is consisted of 89 variables which include 9 growth indices, 30 profitability indices, 23 stability indices, 6 activity indices and 8 productivity indices. For the model building and test, we used 10,925 financial analysis data of total 658 listed firms. PASW Modeler 13 was used to build C5.0 decision trees for the six prediction models. Total 84 variables among financial analysis data are selected as the input variables of each model, and the rights issue status (issued or not issued) is defined as the output variable. To develop prediction models using C5.0 node (Node Options: Output type = Rule set, Use boosting = false, Cross-validate = false, Mode = Simple, Favor = Generality), we used 60% of data for model building and 40% of data for model test. The results of experimental analysis show that the prediction accuracies of data after the IMF financial crisis (59.04% to 60.43%) are about 10 percent higher than ones before IMF financial crisis (68.78% to 71.41%). These results indicate that since the IMF financial crisis, the reliability of financial analysis indices has increased and the firm intention of rights issue has been more obvious. The experiment results also show that the stability-related indices have a major impact on conducting rights issue in the case of short-term prediction. On the other hand, the long-term prediction of conducting rights issue is affected by financial analysis indices on profitability, stability, activity and productivity. All the prediction models include the industry code as one of significant variables. This means that companies in different types of industries show their different types of patterns for rights issue. We conclude that it is desirable for stakeholders to take into account stability-related indices and more various financial analysis indices for short-term prediction and long-term prediction, respectively. The current study has several limitations. First, we need to compare the differences in accuracy by using different data mining techniques such as neural networks, logistic regression and SVM. Second, we are required to develop and to evaluate new prediction models including variables which research in the theory of capital structure has mentioned about the relevance to rights issue.

• Journal of Intelligence and Information Systems
• /
• v.24 no.3
• /
• pp.1-19
• /
• 2018

Large amount of data is now available for research and business sectors to extract knowledge from it. This data can be in the form of unstructured data such as audio, text, and image data and can be analyzed by deep learning methodology. Deep learning is now widely used for various estimation, classification, and prediction problems. Especially, fashion business adopts deep learning techniques for apparel recognition, apparel search and retrieval engine, and automatic product recommendation. The core model of these applications is the image classification using Convolutional Neural Networks (CNN). CNN is made up of neurons which learn parameters such as weights while inputs come through and reach outputs. CNN has layer structure which is best suited for image classification as it is comprised of convolutional layer for generating feature maps, pooling layer for reducing the dimensionality of feature maps, and fully-connected layer for classifying the extracted features. However, most of the classification models have been trained using online product image, which is taken under controlled situation such as apparel image itself or professional model wearing apparel. This image may not be an effective way to train the classification model considering the situation when one might want to classify street fashion image or walking image, which is taken in uncontrolled situation and involves people's movement and unexpected pose. Therefore, we propose to train the model with runway apparel image dataset which captures mobility. This will allow the classification model to be trained with far more variable data and enhance the adaptation with diverse query image. To achieve both convergence and generalization of the model, we apply Transfer Learning on our training network. As Transfer Learning in CNN is composed of pre-training and fine-tuning stages, we divide the training step into two. First, we pre-train our architecture with large-scale dataset, ImageNet dataset, which consists of 1.2 million images with 1000 categories including animals, plants, activities, materials, instrumentations, scenes, and foods. We use GoogLeNet for our main architecture as it has achieved great accuracy with efficiency in ImageNet Large Scale Visual Recognition Challenge (ILSVRC). Second, we fine-tune the network with our own runway image dataset. For the runway image dataset, we could not find any previously and publicly made dataset, so we collect the dataset from Google Image Search attaining 2426 images of 32 major fashion brands including Anna Molinari, Balenciaga, Balmain, Brioni, Burberry, Celine, Chanel, Chloe, Christian Dior, Cividini, Dolce and Gabbana, Emilio Pucci, Ermenegildo, Fendi, Giuliana Teso, Gucci, Issey Miyake, Kenzo, Leonard, Louis Vuitton, Marc Jacobs, Marni, Max Mara, Missoni, Moschino, Ralph Lauren, Roberto Cavalli, Sonia Rykiel, Stella McCartney, Valentino, Versace, and Yve Saint Laurent. We perform 10-folded experiments to consider the random generation of training data, and our proposed model has achieved accuracy of 67.2% on final test. Our research suggests several advantages over previous related studies as to our best knowledge, there haven't been any previous studies which trained the network for apparel image classification based on runway image dataset. We suggest the idea of training model with image capturing all the possible postures, which is denoted as mobility, by using our own runway apparel image dataset. Moreover, by applying Transfer Learning and using checkpoint and parameters provided by Tensorflow Slim, we could save time spent on training the classification model as taking 6 minutes per experiment to train the classifier. This model can be used in many business applications where the query image can be runway image, product image, or street fashion image. To be specific, runway query image can be used for mobile application service during fashion week to facilitate brand search, street style query image can be classified during fashion editorial task to classify and label the brand or style, and website query image can be processed by e-commerce multi-complex service providing item information or recommending similar item.

• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.97-117
• /
• 2020

Product evaluation criteria is an indicator describing attributes or values of products, which enable users or manufacturers measure and understand the products. When companies analyze their products or compare them with competitors, appropriate criteria must be selected for objective evaluation. The criteria should show the features of products that consumers considered when they purchased, used and evaluated the products. However, current evaluation criteria do not reflect different consumers' opinion from product to product. Previous studies tried to used online reviews from e-commerce sites that reflect consumer opinions to extract the features and topics of products and use them as evaluation criteria. However, there is still a limit that they produce irrelevant criteria to products due to extracted or improper words are not refined. To overcome this limitation, this research suggests LDA-k-NN model which extracts possible criteria words from online reviews by using LDA and refines them with k-nearest neighbor. Proposed approach starts with preparation phase, which is constructed with 6 steps. At first, it collects review data from e-commerce websites. Most e-commerce websites classify their selling items by high-level, middle-level, and low-level categories. Review data for preparation phase are gathered from each middle-level category and collapsed later, which is to present single high-level category. Next, nouns, adjectives, adverbs, and verbs are extracted from reviews by getting part of speech information using morpheme analysis module. After preprocessing, words per each topic from review are shown with LDA and only nouns in topic words are chosen as potential words for criteria. Then, words are tagged based on possibility of criteria for each middle-level category. Next, every tagged word is vectorized by pre-trained word embedding model. Finally, k-nearest neighbor case-based approach is used to classify each word with tags. After setting up preparation phase, criteria extraction phase is conducted with low-level categories. This phase starts with crawling reviews in the corresponding low-level category. Same preprocessing as preparation phase is conducted using morpheme analysis module and LDA. Possible criteria words are extracted by getting nouns from the data and vectorized by pre-trained word embedding model. Finally, evaluation criteria are extracted by refining possible criteria words using k-nearest neighbor approach and reference proportion of each word in the words set. To evaluate the performance of the proposed model, an experiment was conducted with review on '11st', one of the biggest e-commerce companies in Korea. Review data were from 'Electronics/Digital' section, one of high-level categories in 11st. For performance evaluation of suggested model, three other models were used for comparing with the suggested model; actual criteria of 11st, a model that extracts nouns by morpheme analysis module and refines them according to word frequency, and a model that extracts nouns from LDA topics and refines them by word frequency. The performance evaluation was set to predict evaluation criteria of 10 low-level categories with the suggested model and 3 models above. Criteria words extracted from each model were combined into a single words set and it was used for survey questionnaires. In the survey, respondents chose every item they consider as appropriate criteria for each category. Each model got its score when chosen words were extracted from that model. The suggested model had higher scores than other models in 8 out of 10 low-level categories. By conducting paired t-tests on scores of each model, we confirmed that the suggested model shows better performance in 26 tests out of 30. In addition, the suggested model was the best model in terms of accuracy. This research proposes evaluation criteria extracting method that combines topic extraction using LDA and refinement with k-nearest neighbor approach. This method overcomes the limits of previous dictionary-based models and frequency-based refinement models. This study can contribute to improve review analysis for deriving business insights in e-commerce market.

• Journal of Intelligence and Information Systems
• /
• v.24 no.2
• /
• pp.125-148
• /
• 2018

Recently, as the demand for big data analysis increases, cases of analyzing unstructured data and using the results are also increasing. Among the various types of unstructured data, text is used as a means of communicating information in almost all fields. In addition, many analysts are interested in the amount of data is very large and relatively easy to collect compared to other unstructured and structured data. Among the various text analysis applications, document classification which classifies documents into predetermined categories, topic modeling which extracts major topics from a large number of documents, sentimental analysis or opinion mining that identifies emotions or opinions contained in texts, and Text Summarization which summarize the main contents from one document or several documents have been actively studied. Especially, the text summarization technique is actively applied in the business through the news summary service, the privacy policy summary service, ect. In addition, much research has been done in academia in accordance with the extraction approach which provides the main elements of the document selectively and the abstraction approach which extracts the elements of the document and composes new sentences by combining them. However, the technique of evaluating the quality of automatically summarized documents has not made much progress compared to the technique of automatic text summarization. Most of existing studies dealing with the quality evaluation of summarization were carried out manual summarization of document, using them as reference documents, and measuring the similarity between the automatic summary and reference document. Specifically, automatic summarization is performed through various techniques from full text, and comparison with reference document, which is an ideal summary document, is performed for measuring the quality of automatic summarization. Reference documents are provided in two major ways, the most common way is manual summarization, in which a person creates an ideal summary by hand. Since this method requires human intervention in the process of preparing the summary, it takes a lot of time and cost to write the summary, and there is a limitation that the evaluation result may be different depending on the subject of the summarizer. Therefore, in order to overcome these limitations, attempts have been made to measure the quality of summary documents without human intervention. On the other hand, as a representative attempt to overcome these limitations, a method has been recently devised to reduce the size of the full text and to measure the similarity of the reduced full text and the automatic summary. In this method, the more frequent term in the full text appears in the summary, the better the quality of the summary. However, since summarization essentially means minimizing a lot of content while minimizing content omissions, it is unreasonable to say that a "good summary" based on only frequency always means a "good summary" in its essential meaning. In order to overcome the limitations of this previous study of summarization evaluation, this study proposes an automatic quality evaluation for text summarization method based on the essential meaning of summarization. Specifically, the concept of succinctness is defined as an element indicating how few duplicated contents among the sentences of the summary, and completeness is defined as an element that indicating how few of the contents are not included in the summary. In this paper, we propose a method for automatic quality evaluation of text summarization based on the concepts of succinctness and completeness. In order to evaluate the practical applicability of the proposed methodology, 29,671 sentences were extracted from TripAdvisor 's hotel reviews, summarized the reviews by each hotel and presented the results of the experiments conducted on evaluation of the quality of summaries in accordance to the proposed methodology. It also provides a way to integrate the completeness and succinctness in the trade-off relationship into the F-Score, and propose a method to perform the optimal summarization by changing the threshold of the sentence similarity.