• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.1-17
• /
• 2017

The deep learning framework is software designed to help develop deep learning models. Some of its important functions include "automatic differentiation" and "utilization of GPU". The list of popular deep learning framework includes Caffe (BVLC) and Theano (University of Montreal). And recently, Microsoft's deep learning framework, Microsoft Cognitive Toolkit, was released as open-source license, following Google's Tensorflow a year earlier. The early deep learning frameworks have been developed mainly for research at universities. Beginning with the inception of Tensorflow, however, it seems that companies such as Microsoft and Facebook have started to join the competition of framework development. Given the trend, Google and other companies are expected to continue investing in the deep learning framework to bring forward the initiative in the artificial intelligence business. From this point of view, we think it is a good time to compare some of deep learning frameworks. So we compare three deep learning frameworks which can be used as a Python library. Those are Google's Tensorflow, Microsoft's CNTK, and Theano which is sort of a predecessor of the preceding two. The most common and important function of deep learning frameworks is the ability to perform automatic differentiation. Basically all the mathematical expressions of deep learning models can be represented as computational graphs, which consist of nodes and edges. Partial derivatives on each edge of a computational graph can then be obtained. With the partial derivatives, we can let software compute differentiation of any node with respect to any variable by utilizing chain rule of Calculus. First of all, the convenience of coding is in the order of CNTK, Tensorflow, and Theano. The criterion is simply based on the lengths of the codes and the learning curve and the ease of coding are not the main concern. According to the criteria, Theano was the most difficult to implement with, and CNTK and Tensorflow were somewhat easier. With Tensorflow, we need to define weight variables and biases explicitly. The reason that CNTK and Tensorflow are easier to implement with is that those frameworks provide us with more abstraction than Theano. We, however, need to mention that low-level coding is not always bad. It gives us flexibility of coding. With the low-level coding such as in Theano, we can implement and test any new deep learning models or any new search methods that we can think of. The assessment of the execution speed of each framework is that there is not meaningful difference. According to the experiment, execution speeds of Theano and Tensorflow are very similar, although the experiment was limited to a CNN model. In the case of CNTK, the experimental environment was not maintained as the same. The code written in CNTK has to be run in PC environment without GPU where codes execute as much as 50 times slower than with GPU. But we concluded that the difference of execution speed was within the range of variation caused by the different hardware setup. In this study, we compared three types of deep learning framework: Theano, Tensorflow, and CNTK. According to Wikipedia, there are 12 available deep learning frameworks. And 15 different attributes differentiate each framework. Some of the important attributes would include interface language (Python, C ++, Java, etc.) and the availability of libraries on various deep learning models such as CNN, RNN, DBN, and etc. And if a user implements a large scale deep learning model, it will also be important to support multiple GPU or multiple servers. Also, if you are learning the deep learning model, it would also be important if there are enough examples and references.

• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.39-70
• /
• 2017

Data-drive analytics techniques have been recently applied to public surveys. Instead of simply gathering survey results or expert opinions to research the preference for a recently launched product, enterprises need a way to collect and analyze various types of online data and then accurately figure out customer preferences. In the main concept of existing data-based survey methods, the sentiment lexicon for a particular domain is first constructed by domain experts who usually judge the positive, neutral, or negative meanings of the frequently used words from the collected text documents. In order to research the preference for a particular product, the existing approach collects (1) review posts, which are related to the product, from several product review web sites; (2) extracts sentences (or phrases) in the collection after the pre-processing step such as stemming and removal of stop words is performed; (3) classifies the polarity (either positive or negative sense) of each sentence (or phrase) based on the sentiment lexicon; and (4) estimates the positive and negative ratios of the product by dividing the total numbers of the positive and negative sentences (or phrases) by the total number of the sentences (or phrases) in the collection. Furthermore, the existing approach automatically finds important sentences (or phrases) including the positive and negative meaning to/against the product. As a motivated example, given a product like Sonata made by Hyundai Motors, customers often want to see the summary note including what positive points are in the 'car design' aspect as well as what negative points are in thesame aspect. They also want to gain more useful information regarding other aspects such as 'car quality', 'car performance', and 'car service.' Such an information will enable customers to make good choice when they attempt to purchase brand-new vehicles. In addition, automobile makers will be able to figure out the preference and positive/negative points for new models on market. In the near future, the weak points of the models will be improved by the sentiment analysis. For this, the existing approach computes the sentiment score of each sentence (or phrase) and then selects top-k sentences (or phrases) with the highest positive and negative scores. However, the existing approach has several shortcomings and is limited to apply to real applications. The main disadvantages of the existing approach is as follows: (1) The main aspects (e.g., car design, quality, performance, and service) to a product (e.g., Hyundai Sonata) are not considered. Through the sentiment analysis without considering aspects, as a result, the summary note including the positive and negative ratios of the product and top-k sentences (or phrases) with the highest sentiment scores in the entire corpus is just reported to customers and car makers. This approach is not enough and main aspects of the target product need to be considered in the sentiment analysis. (2) In general, since the same word has different meanings across different domains, the sentiment lexicon which is proper to each domain needs to be constructed. The efficient way to construct the sentiment lexicon per domain is required because the sentiment lexicon construction is labor intensive and time consuming. To address the above problems, in this article, we propose a novel product reputation mining algorithm that (1) extracts topics hidden in review documents written by customers; (2) mines main aspects based on the extracted topics; (3) measures the positive and negative ratios of the product using the aspects; and (4) presents the digest in which a few important sentences with the positive and negative meanings are listed in each aspect. Unlike the existing approach, using hidden topics makes experts construct the sentimental lexicon easily and quickly. Furthermore, reinforcing topic semantics, we can improve the accuracy of the product reputation mining algorithms more largely than that of the existing approach. In the experiments, we collected large review documents to the domestic vehicles such as K5, SM5, and Avante; measured the positive and negative ratios of the three cars; showed top-k positive and negative summaries per aspect; and conducted statistical analysis. Our experimental results clearly show the effectiveness of the proposed method, compared with the existing method.

• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.139-160
• /
• 2017

Technology commercialization creates effective economic value by linking the company's R & D processes and outputs to the market. This technology commercialization is important in that a company can retain and maintain a sustained competitive advantage. In order for a specific technology to be commercialized, it goes through the stage of technical planning, technology research and development, and commercialization. This process involves a lot of time and money. Therefore, the duration and cost of technology commercialization are important decision information for determining the market entry strategy. In addition, it is more important information for a technology investor to rationally evaluate the technology value. In this way, it is very important to scientifically estimate the duration and cost of the technology commercialization. However, research on technology commercialization is insufficient and related methodology are lacking. In this study, we propose an evaluation model that can estimate the duration and cost of R & D technology commercialization for small and medium-sized enterprises. To accomplish this, this study collected the public data of the National Science & Technology Information Service (NTIS) and the survey data provided by the Small and Medium Business Administration. Also this study will develop the estimation model of commercialization duration and cost of R&D performance on using these data based on the market approach, one of the technology valuation methods. Specifically, this study defined the process of commercialization as consisting of development planning, development progress, and commercialization. We collected the data from the NTIS database and the survey of SMEs technical statistics of the Small and Medium Business Administration. We derived the key variables such as stage-wise R&D costs and duration, the factors of the technology itself, the factors of the technology development, and the environmental factors. At first, given data, we estimates the costs and duration in each technology readiness level (basic research, applied research, development research, prototype production, commercialization), for each industry classification. Then, we developed and verified the research model of each industry classification. The results of this study can be summarized as follows. Firstly, it is reflected in the technology valuation model and can be used to estimate the objective economic value of technology. The duration and the cost from the technology development stage to the commercialization stage is a critical factor that has a great influence on the amount of money to discount the future sales from the technology. The results of this study can contribute to more reliable technology valuation because it estimates the commercialization duration and cost scientifically based on past data. Secondly, we have verified models of various fields such as statistical model and data mining model. The statistical model helps us to find the important factors to estimate the duration and cost of technology Commercialization, and the data mining model gives us the rules or algorithms to be applied to an advanced technology valuation system. Finally, this study reaffirms the importance of commercialization costs and durations, which has not been actively studied in previous studies. The results confirm the significant factors to affect the commercialization costs and duration, furthermore the factors are different depending on industry classification. Practically, the results of this study can be reflected in the technology valuation system, which can be provided by national research institutes and R & D staff to provide sophisticated technology valuation. The relevant logic or algorithm of the research result can be implemented independently so that it can be directly reflected in the system, so researchers can use it practically immediately. In conclusion, the results of this study can be a great contribution not only to the theoretical contributions but also to the practical ones.

• Journal of Intelligence and Information Systems
• /
• v.21 no.4
• /
• pp.1-16
• /
• 2015

Traffic accident is one of the major cause of death worldwide for the last several decades. According to the statistics of world health organization, approximately 1.24 million deaths occurred on the world's roads in 2010. In order to reduce future traffic accident, multipronged approaches have been adopted including traffic regulations, injury-reducing technologies, driving training program and so on. Records on traffic accidents are generated and maintained for this purpose. To make these records meaningful and effective, it is necessary to analyze relationship between traffic accident and related factors including vehicle design, road design, weather, driver behavior etc. Insight derived from these analysis can be used for accident prevention approaches. Traffic accident data mining is an activity to find useful knowledges about such relationship that is not well-known and user may interested in it. Many studies about mining accident data have been reported over the past two decades. Most of studies mainly focused on predict risk of accident using accident related factors. Supervised learning methods like decision tree, logistic regression, k-nearest neighbor, neural network are used for these prediction. However, derived prediction model from these algorithms are too complex to understand for human itself because the main purpose of these algorithms are prediction, not explanation of the data. Some of studies use unsupervised clustering algorithm to dividing the data into several groups, but derived group itself is still not easy to understand for human, so it is necessary to do some additional analytic works. Rule based learning methods are adequate when we want to derive comprehensive form of knowledge about the target domain. It derives a set of if-then rules that represent relationship between the target feature with other features. Rules are fairly easy for human to understand its meaning therefore it can help provide insight and comprehensible results for human. Association rule learning methods and subgroup discovery methods are representing rule based learning methods for descriptive task. These two algorithms have been used in a wide range of area from transaction analysis, accident data analysis, detection of statistically significant patient risk groups, discovering key person in social communities and so on. We use both the association rule learning method and the subgroup discovery method to discover useful patterns from a traffic accident dataset consisting of many features including profile of driver, location of accident, types of accident, information of vehicle, violation of regulation and so on. The association rule learning method, which is one of the unsupervised learning methods, searches for frequent item sets from the data and translates them into rules. In contrast, the subgroup discovery method is a kind of supervised learning method that discovers rules of user specified concepts satisfying certain degree of generality and unusualness. Depending on what aspect of the data we are focusing our attention to, we may combine different multiple relevant features of interest to make a synthetic target feature, and give it to the rule learning algorithms. After a set of rules is derived, some postprocessing steps are taken to make the ruleset more compact and easier to understand by removing some uninteresting or redundant rules. We conducted a set of experiments of mining our traffic accident data in both unsupervised mode and supervised mode for comparison of these rule based learning algorithms. Experiments with the traffic accident data reveals that the association rule learning, in its pure unsupervised mode, can discover some hidden relationship among the features. Under supervised learning setting with combinatorial target feature, however, the subgroup discovery method finds good rules much more easily than the association rule learning method that requires a lot of efforts to tune the parameters.

• Journal of Intelligence and Information Systems
• /
• v.12 no.1
• /
• pp.17-31
• /
• 2006

In this paper, passport recognition and face verification methods which can automatically recognize passport codes and discriminate forgery passports to improve efficiency and systematic control of immigration management are proposed. Adjusting the slant is very important for recognition of characters and face verification since slanted passport images can bring various unwanted effects to the recognition of individual codes and faces. Therefore, after smearing the passport image, the longest extracted string of characters is selected. The angle adjustment can be conducted by using the slant of the straight and horizontal line that connects the center of thickness between left and right parts of the string. Extracting passport codes is done by Sobel operator, horizontal smearing, and 8-neighborhood contour tracking algorithm. The string of codes can be transformed into binary format by applying repeating binary method to the area of the extracted passport code strings. The string codes are restored by applying CDM mask to the binary string area and individual codes are extracted by 8-neighborhood contour tracking algerian. The proposed RBF network is applied to the middle layer of RBF network by using the fuzzy logic connection operator and proposing the enhanced fuzzy ART algorithm that dynamically controls the vigilance parameter. The face is authenticated by measuring the similarity between the feature vector of the facial image from the passport and feature vector of the facial image from the database that is constructed with PCA algorithm. After several tests using a forged passport and the passport with slanted images, the proposed method was proven to be effective in recognizing passport codes and verifying facial images.

• Journal of Intelligence and Information Systems
• /
• v.12 no.1
• /
• pp.1-15
• /
• 2006

In Korea, a resident registration card has various personal information such as a present address, a resident registration number, a face picture and a fingerprint. A plastic-type resident card currently used is easy to forge or alter and tricks of forgery grow to be high-degree as time goes on. So, whether a resident card is forged or not is difficult to judge by only an examination with the naked eye. This paper proposed an automatic recognition method of a resident card which recognizes a resident registration number by using a refined ART2-based RBF network newly proposed and authenticates a face picture by a template image matching method. The proposed method, first, extracts areas including a resident registration number and the date of issue from a resident card image by applying Sobel masking, median filtering and horizontal smearing operations to the image in turn. To improve the extraction of individual codes from extracted areas, the original image is binarized by using a high-frequency passing filter and CDM masking is applied to the binaried image fur making image information of individual codes better. Lastly, individual codes, which are targets of recognition, are extracted by applying 4-directional contour tracking algorithm to extracted areas in the binarized image. And this paper proposed a refined ART2-based RBF network to recognize individual codes, which applies ART2 as the loaming structure of the middle layer and dynamicaly adjusts a teaming rate in the teaming of the middle and the output layers by using a fuzzy control method to improve the performance of teaming. Also, for the precise judgement of forgey of a resident card, the proposed method supports a face authentication by using a face template database and a template image matching method. For performance evaluation of the proposed method, this paper maked metamorphoses of an original image of resident card such as a forgey of face picture, an addition of noise, variations of contrast variations of intensity and image blurring, and applied these images with original images to experiments. The results of experiment showed that the proposed method is excellent in the recognition of individual codes and the face authentication fur the automatic recognition of a resident card.

• Journal of Intelligence and Information Systems
• /
• v.17 no.1
• /
• pp.71-90
• /
• 2011

This study was designed to empirically analyze the effect of control activities(physical, managerial and technical securities) of information protection on organizational effectiveness and the mediating effects of information application. The result was summarized as follows. First, the effect of control activities(physical, technical and managerial securities) of information protection on organizational effectiveness showed that the physical, technical and managerial security factors have a significant positive effect on the organizational effectiveness(p < .01). Second, the effect of control activities(physical, technical and managerial securities) of information protection on information application showed that the technical and managerial security factors have a significant positive effect on the information application(p < .01). Third, the explanatory power of models, which additionally put the information protection control activities(physical, technical and managerial securities) and the interaction variables of information application to verify how the information protection control activities( physical, technical and managerial security controls) affecting the organizational effectiveness are mediated by the information application, was 50.6%~4.1% additional increase. And the interaction factor(${\beta}$ = .148, p < .01) of physical security and information application, and interaction factor(${\beta}$ = .196, p < .01) of physical security and information application among additionally-put interaction variables, were statistically significant(p < .01), indicating the information application has mediated the relationship between physical security and managerial security factors of control activities, and organizational effectiveness. As for results stated above, it was proven that physical, technical and managerial factors as internal control activities for information protection are main mechanisms affecting the organizational effectiveness very significantly by information application. In information protection control activities, the more all physical, technical and managerial security factors were efficiently well performed, the higher information application, and the more information application was efficiently controlled and mediated, which it was proven that all these three factors are variables for useful information application. It suggested that they have acted as promotion mechanisms showing a very significant result on the internal customer satisfaction of employees, the efficiency of information management and the reduction of risk in the organizational effectiveness for information protection by the mediating or difficulty of proved information application.

• Journal of Intelligence and Information Systems
• /
• v.17 no.1
• /
• pp.111-125
• /
• 2011

Clustering is a process of grouping similar or relevant documents into a cluster and assigning a meaningful concept to the cluster. By this process, clustering facilitates fast and correct search for the relevant documents by narrowing down the range of searching only to the collection of documents belonging to related clusters. For effective clustering, techniques are required for identifying similar documents and grouping them into a cluster, and discovering a concept that is most relevant to the cluster. One of the problems often appearing in this context is the detection of a complex concept that overlaps with several simple concepts at the same hierarchical level. Previous clustering methods were unable to identify and represent a complex concept that belongs to several different clusters at the same level in the concept hierarchy, and also could not validate the semantic hierarchical relationship between a complex concept and each of simple concepts. In order to solve these problems, this paper proposes a new clustering method that identifies and represents complex concepts efficiently. We developed the Hierarchical Overlapping Clustering (HOC) algorithm that modified the traditional Agglomerative Hierarchical Clustering algorithm to allow overlapped clusters at the same level in the concept hierarchy. The HOC algorithm represents the clustering result not by a tree but by a lattice to detect complex concepts. We developed a system that employs the HOC algorithm to carry out the goal of complex concept detection. This system operates in three phases; 1) the preprocessing of documents, 2) the clustering using the HOC algorithm, and 3) the validation of semantic hierarchical relationships among the concepts in the lattice obtained as a result of clustering. The preprocessing phase represents the documents as x-y coordinate values in a 2-dimensional space by considering the weights of terms appearing in the documents. First, it goes through some refinement process by applying stopwords removal and stemming to extract index terms. Then, each index term is assigned a TF-IDF weight value and the x-y coordinate value for each document is determined by combining the TF-IDF values of the terms in it. The clustering phase uses the HOC algorithm in which the similarity between the documents is calculated by applying the Euclidean distance method. Initially, a cluster is generated for each document by grouping those documents that are closest to it. Then, the distance between any two clusters is measured, grouping the closest clusters as a new cluster. This process is repeated until the root cluster is generated. In the validation phase, the feature selection method is applied to validate the appropriateness of the cluster concepts built by the HOC algorithm to see if they have meaningful hierarchical relationships. Feature selection is a method of extracting key features from a document by identifying and assigning weight values to important and representative terms in the document. In order to correctly select key features, a method is needed to determine how each term contributes to the class of the document. Among several methods achieving this goal, this paper adopted the $x^2$�� statistics, which measures the dependency degree of a term t to a class c, and represents the relationship between t and c by a numerical value. To demonstrate the effectiveness of the HOC algorithm, a series of performance evaluation is carried out by using a well-known Reuter-21578 news collection. The result of performance evaluation showed that the HOC algorithm greatly contributes to detecting and producing complex concepts by generating the concept hierarchy in a lattice structure.

• Journal of Intelligence and Information Systems
• /
• v.17 no.1
• /
• pp.91-110
• /
• 2011

This paper analyzes the cases where the hotels have increased their services and enhanced their work process through IT solutions to cope with computerization globalization. Also the cases have been studies where national hotels use the CRM solution internally to respond effectively to customers requests, increase customer analysis, and build marketing strategies. In particular, this study discusses the introduction of the CRM solutions and CRM sales business and marketing services using a process for utilizing the presumed, CRM by introducing effective DEA(Data Envelopment Analysis). First, the comparison has done regarding the relative efficiency of L Company with the CCR model, then compared L Company's restaurants and facilities' effectiveness through BCC model. L Company reached a conclusion that it is important to precisely create and manage sales data which are the preliminary data for CRM, and for that reason it made it possible to save sales data generated by POS system on each sales performance database. In order to do that, it newly established Oracle POS system and LORIS POS system concerned with restaurants for food and beverage as well as rooms, and made it possible to stably generate and manage sales data and manage. Moreover, it set up a composite database to control comprehensively the results of work processes during a specific period by collecting customer registration information and made it possible to systematically control the information on sales performances. By establishing a system which unifies database and managing it comprehensively, impeccability of data has been greatly enhanced and a problem which generated asymmetric data could be thoroughly solved. Using data accumulated on the comprehensive database, sales data can be analyzed, categorized, classified through data mining engine imbedded in Polaris CRM and the results can be organized on data mart to provide them in the form of CRM application data. By transforming original sales data into forms which are easy to handle and saving them on data mart separately, it enabled acquiring well-organized data with ease when engaging in various marketing operations, holding a morning meeting and working on decision-making. By using summarized data at data mart, it was possible to process marketing operations such as telemarketing, direct mailing, internet marketing service and service product developments for perceived customers; moreover, information on customer perceptions which is one of CRM's end-products could feed back into the comprehensive database. This research was undertaken to find out how effectively CRM has been employed by comparing and analyzing the management performance of each enterprise site and store after introducing CRM to Hotel enterprises using DEA technique. According to the research results, efficiency evaluation for each site was calculated through input and output factors to find out comparative CRM system usage efficiency of L's Company four sites; moreover, with regard to stores, the sizes of workforce and budget application show a huge difference and so does the each store efficiency. Furthermore, by using the DEA technique, it could assess which sites have comparatively high efficiency and which don't by comparing and evaluating hotel enterprises IT project outcomes such as CRM introduction using the CCR model for each site of the related enterprises. By using the BCC model, it could comparatively evaluate the outcome of CRM usage at each store of A site, which is representative of L Company, and as a result, it could figure out which stores maintain high efficiency in using CRM and which don't. It analyzed the cases of CRM introduction at L Company, which is a hotel enterprise, and precisely evaluated them through DEA. L Company analyzed the customer analysis system by introducing CRM and achieved to provide customers identified through client analysis data with one to one tailored services. Moreover, it could come up with a plan to differentiate the service for customers who revisit by assessing customer discernment rate. As tasks to be solved in the future, it is required to do research on the process analysis which can lead to a specific outcome such as increased sales volumes by carrying on test marketing, target marketing using CRM. Furthermore, it is also necessary to do research on efficiency evaluation in accordance with linkages between other IT solutions such as ERP and CRM system.

• Journal of Intelligence and Information Systems
• /
• v.17 no.1
• /
• pp.127-138
• /
• 2011

The core service of most research portal sites is providing relevant research papers to various researchers that match their research interests. This kind of service may only be effective and easy to use when a user can provide correct and concrete information about a paper such as the title, authors, and keywords. However, unfortunately, most users of this service are not acquainted with concrete bibliographic information. It implies that most users inevitably experience repeated trial and error attempts of keyword-based search. Especially, retrieving a relevant research paper is more difficult when a user is novice in the research domain and does not know appropriate keywords. In this case, a user should perform iterative searches as follows : i) perform an initial search with an arbitrary keyword, ii) acquire related keywords from the retrieved papers, and iii) perform another search again with the acquired keywords. This usage pattern implies that the level of service quality and user satisfaction of a portal site are strongly affected by the level of keyword management and searching mechanism. To overcome this kind of inefficiency, some leading research portal sites adopt the association rule mining-based keyword recommendation service that is similar to the product recommendation of online shopping malls. However, keyword recommendation only based on association analysis has limitation that it can show only a simple and direct relationship between two keywords. In other words, the association analysis itself is unable to present the complex relationships among many keywords in some adjacent research areas. To overcome this limitation, we propose the hybrid approach for establishing association network among keywords used in research papers. The keyword association network can be established by the following phases : i) a set of keywords specified in a certain paper are regarded as co-purchased items, ii) perform association analysis for the keywords and extract frequent patterns of keywords that satisfy predefined thresholds of confidence, support, and lift, and iii) schematize the frequent keyword patterns as a network to show the core keywords of each research area and connecting keywords among two or more research areas. To estimate the practical application of our approach, we performed a simple experiment with 600 keywords. The keywords are extracted from 131 research papers published in five prominent Korean journals in 2009. In the experiment, we used the SAS Enterprise Miner for association analysis and the R software for social network analysis. As the final outcome, we presented a network diagram and a cluster dendrogram for the keyword association network. We summarized the results in Section 4 of this paper. The main contribution of our proposed approach can be found in the following aspects : i) the keyword network can provide an initial roadmap of a research area to researchers who are novice in the domain, ii) a researcher can grasp the distribution of many keywords neighboring to a certain keyword, and iii) researchers can get some idea for converging different research areas by observing connecting keywords in the keyword association network. Further studies should include the following. First, the current version of our approach does not implement a standard meta-dictionary. For practical use, homonyms, synonyms, and multilingual problems should be resolved with a standard meta-dictionary. Additionally, more clear guidelines for clustering research areas and defining core and connecting keywords should be provided. Finally, intensive experiments not only on Korean research papers but also on international papers should be performed in further studies.