유사도를 이용한 오프라인 상품 기반 추천 기법

김철진,조천우,정지현 한국지식정보기술학회 2019 한국지식정보기술학회 논문지 Vol.14 No.4

The online shopping mall can efficiently provide the recommendation product to the user by using the purchase transaction information of the user. However, in the offline store, there is a limit in providing recommended products in real time using information of users or purchase transaction information. Currently, O2O service provision is spreading, but development and research on personalized recommendation service based on offline products are insufficient. In this paper, we propose an architecture for recommending products suitable for users by calculating similarity between products based on offline individual products and online transaction information. We also propose a procedure for deriving a recommendation product among the constituent modules constituting the architecture. The offline individual product is identified through the Beacon sensor, and the user selects the offline product received from the beacon sensor to determine interest. It calculates the similarity based on offline products and online transaction information and provides top-n recommended products to users. We prove the feasibility of the architecture of this study by constructing a system that recommends products that interest the user by calculating the similarity for offline clothing of clothing store. The existing researches recommend brand based on the purchase history of the offline store visited by the user, but in this paper, it is different in terms of providing recommended products for individual products.

데이터 분석 기반 추천학습모델 생성도구 설계 및 구현

정지현,조천우,변동훈,김철진 한국지식정보기술학회 2021 한국지식정보기술학회 논문지 Vol.16 No.1

Current sales of offline products are conducted on a face-to-face recommendation based on customer’s preferences and fields of interest. However, as the online shopping malls are actively spreading to the public, recommendation researches are ongoing to make more accurate recommendation using customers’ online activity data besides a face-to-face recommendation. But compared to recommendation systems of large-scale shopping mall where operates big-scale database and professional human resources, small-scale shopping malls use simple recommenders relatively, which can be interpreted as a judgement that those recommenders developed by large-scale companies are unfit to apply onto their system. In this paper, we propose a platform that can develop a generalized recommendation tool by utilizing Open API. Recommender-development tool of this study refines datasets to be used for developing a recommendation system, analyzes the datasets to derive proper algorithms. After doing those progresses, it proceeds model learning progress based on the derived algorithm and provides Open API to a user so that the user can apply the recommendation system to own shopping mall system. We prove proposed tool by developing recommendation system using real transaction datasets with this tool, lastly mounting the Open API onto the test shopping mall site. While users using existing tools must be directly specified the schema of datasets to make a recommendation system, this study has a distinction in that it allows users to develop the recommendation model by automatically proceeding with this process.

상품 추천 기법의 성능평가 분석

김철진,정지현,조천우,유제광 한국지식정보기술학회 2019 한국지식정보기술학회 논문지 Vol.14 No.5

In the vast product information of the e-commerce, the user needs a lot of effort to find the required product, and the seller may affect the sales if the product is not provided quickly to the user. Accordingly, the e-commerce company provides a recommendation service based on the user's past purchase information so that the user can provide a product required by the user. The recommendation techniques for providing recommendation services include a collaborative filtering recommendation technique that derives recommendation information through a relationship between users or products and a recommendation technique that utilizes a deep learning technology based on machine learning. In this paper, we study user-based collaborative filtering recommendation and item-based collaborative filtering recommendation as collaborative filtering recommendation techniques, and RNN, LSTM, and Word2Vec recommendation techniques as deep recommendation techniques. In this paper, we evaluate the recommendation performance based on the e-commerce purchase information for these recommendation techniques. As a metric for evaluating the recommendation performance, it analyzes the recommendation performance using accuracy, recall, and F1 measure. The results of the validation of the recommendation performance showed that the LSTM recommendation technique had the best recommendation performance, and that the recommendation performance was the best when the number of recommendations was Top-10. Based on the recommended performance evaluation procedure and evaluation results proposed in this study, it can be referred to when analyzing the performance of recommendations in various fields.

전자상거래 추천 모델 개발 플랫폼을 위한 오픈 아키텍쳐 및 Open API

김철진,정지현,조천우,변동훈 한국지식정보기술학회 2020 한국지식정보기술학회 논문지 Vol.15 No.6

e-Commerce recommendation service is an essential function and plays an important role in increasing sales since it is provided in connection with functions such as product search, order processing, and shopping cart. This recommendation service requires a high level of technology from developers developing e-commerce, so it is developed by a specific artificial intelligence engineer or applied by introducing an external recommendation solution. Integration of recommended services by external solutions or external development companies cannot satisfy the requirements of e-commerce services to be developed, and cannot provide rapid maintenance due to frequent data changes. Accordingly, research on a generalized development platform for generalizing and providing recommendation services suitable for a specific domain or developing a recommendation service is being actively conducted. Amazon Personalize service and Microsoft Azure Machine Learning service are generalized tools for developing recommended services by developers. However, these recommendation model development tools have a workload of defining essential data information for training data required to generate a recommendation model. In this paper, we derive a learning algorithm without defining data by using an association analysis algorithm between data for analyzing learning data. Also, based on the derived learning algorithm, we propose an Open API for developing and verifying a recommendation model. In the experiment, the learning algorithm is derived and the open API is verified by using the open transaction data of the e-commerce transaction. Through this, the suitability of the open architecture of the recommendation model development platform is verified.

상품 추천 정확도를 향상시키기 위한 점증적 추천 기법

김철진,정지현,조천우 한국지식정보기술학회 2021 한국지식정보기술학회 논문지 Vol.16 No.3

In order to recommend products of interest in existing e-commerce transactions, recommended products are derived through CRM or big data analysis, or products are recommended using a recommendation technique applied with machine learning technology. Among the existing recommendation techniques, a recommendation technique using a collaborative filtering technique or machine learning technique has had a great influence on improving the purchasing power of a product to users. However, the purchase data generated by the recommendation data provided to the customer by the recommendation model generated through various recommendation techniques is meaningful as data for generating more reliable recommendations for purchases that will occur in the future, but was not used. Therefore, in this research, in order to improve the accuracy of e-commerce recommendation, we use customer purchase data generated through data for product recommendation, and propose an incremental recommendation technique based on this. Based on the LSTM model, we propose the architecture and procedure of the incremental recommendation technique. In the experiment, to verify the proposed architecture, we learn using the published e-commerce data and verify the accuracy through the separated data. Also, the accuracy of the technique proposed in this study is compared and analyzed with the existing recommended techniques.

D.I.Y : 머신러닝 교육을 위한 블록 기반 프로그래밍 플랫폼

이세훈(Se-hoon Lee),정지현(Ji-hyun Jeong),이진형(Jin-hyeong Lee),조천우(Cheon-woo Jo) 한국컴퓨터정보학회 2020 한국컴퓨터정보학회 학술발표논문집 Vol.28 No.2