Empirical Comparison of the Effects of Online and Offline Recommendation Duration on Purchasing Decisions: Case of Korea Food E-commerce Company

LI QINGLONG,정재호,Dongeon Kim,이흠철,Ilyoung Choi,김재경 한국경영정보학회 2024 Asia Pacific Journal of Information Systems Vol.34 No.1

Most studies on recommender systems to evaluate recommendation performances focus on offline evaluation methods utilizing past customer transaction records. However, evaluating recommendation performance through real-world stimulation becomes challenging. Moreover, such methods cannot evaluate the duration of the recommendation effect. This study measures the personalized recommendation (stimulus) effect when the product recommendation to customers leads to actual purchases and evaluates the duration of the stimulus personalized recommendation effect leading to purchases. The results revealed a 4.58% improvement in recommendation performance in the online environment compared with that in the offline environment. Furthermore, there is little difference in recommendation performance in offline experiments by period, whereas the recommendation performance declines with time in online experiments.

컨텍스트 정보의 속성 중요도를 고려한 딥러닝 기반 추천 모델 성능 향상에 관한 연구

LI QINGLONG,장동수,구하은,김재경 한국지능정보시스템학회 2024 지능정보연구 Vol.30 No.1

추천 시스템의 데이터 희소성 문제를 개선하기 위해 컨텍스트 정보를 활용한 연구가 활발하게 진행되고 있다. 다양한 선행 연구들은 추천 모델을 구축하는 과정에서 컨텍스트 정보를 구성하는 속성들을 단순히 결합하는 방법을 활용하였다. 하지만 제품에 대한 선호도는 고객이 고려하는 속성의 중요도에 따라 달라질 수 있기 때문에 컨텍스트 정보를 단순히 결합하는 방법은 추천 모델에 효과적이지 않을 수 있다. 따라서 본 연구는 고객과 제품 간 상호작용을 정교하게 학습하기 위해 컨텍스트 정보를 구성하는 각 속성의 중요도를 반영한 CINCF(Contextual information Importance-based Neural Collaborative Filtering) 모델을 제안한다. CINCF는 다양한 컨텍스트 정보에 포함된 속성들을 통합하고 어텐션 메커니즘을 활용하여 속성들이 가질 중요도를 고려하여 상호작용을 학습한다. 본 연구에서 제안된 CINCF의 추천 성능을 검증하기 위해 Amazon.com에서 수집한 세 가지 카테고리의 데이터를 사용하였다. CINCF 모델은 여러 베이스라인 모델과 비교하여 우수한 추천 성능을 보였으며, 이를 통해 컨텍스트 정보에 내포된 속성의 중요도를 고려하는 추천 방법론의 효과성을 입증할 수 있었다. Recommendation studies using context information are actively conducted to improve the data sparsity problem of the recommender system. Previous studies used simply combining attributes containing context information to construct a recommendation model. However, simply combining contextual information may not be effective for a recommendation model because the customer’s preference for a product may vary depending on the importance of attributes. This study proposes the CINCF (Contextual information Importance-based Neural Collaborative Filtering) model that reflects the importance of each attribute containing context information to learn the interaction between the customer and the product elaborately. CINCF integrates attributes in various contexts information and learns interactions by considering the importance of attributes through an attention mechanism. To evaluate the recommendation performance of the proposed model, three categories of data collected by Amazon.com were used. Compared to baseline models, the CINCF model showed excellent recommendation performance. This shows the effectiveness of the recommendation methodology that considers the importance of attributes containing context information.

CNN 기반 리뷰 유용성 점수 예측을 통한 개인화 추천 서비스 성능 향상에 관한 연구

이청용(Qinglong Li),이병현(Byunghyun Lee),이흠철(Xinzhe Li),김재경(Jae Kyeong Kim) 한국지능정보시스템학회 2021 지능정보연구 Vol.27 No.3

전자상거래 시장이 빠르게 성장하면서 다양한 유형의 제품이 출시되고 있으며, 이로 인해 사용자들은 구매 의사결정과정에 많은 시간이 소요되는 정보 과부하 문제에 직면하고 있다. 따라서 사용자에게 맞춤형 제품 및 서비스를 제공해줄 수 있는 개인화 추천 서비스의 중요성이 대두되고 있다. 대표적으로 Netflix, Amazon, Google 등 세계적 기업은 개인화 추천 서비스를 도입하여 사용자의 구매 의사결정을 지원하고 있다. 이에 따라 사용자의 정보탐색 비용이 감소하는 효과가 나타났고, 기업의 매출 상승에도 긍정적인 영향을 끼치고 있다. 기존 개인화 추천 서비스 관련 연구에서 주로 사용된 협업 필터링(Collaborative Filtering, CF) 기법은 정량화된 정보를 활용하여 사용자의 선호도를 예측하였다. 그러나 정량화된 정보만을 활용하면 사용자의 구매 의도는 고려하지 못하므로 추천 성능이 저하될 수 있다는 문제점이 제기되고 있다. 이와 같은 기존 연구의 문제점을 개선하기 위해 최근에는 사용자가 작성한 리뷰를 활용한 개인화 추천 서비스 연구가 활발히 진행되고 있다. 그러나 리뷰에는 광고성 내용, 거짓 후기, 의미를 전혀 파악할 수 없거나 제품과 관련 없는 내용 등 구매 의사결정을 저해하는 요소들이 포함되어 있다. 이러한 요소들이 포함된 리뷰를 활용하여 추천 서비스를 제공하게 되면, 추천 성능이 저하되는 문제가 발생할 수 있다. 따라서 본 연구에서는 이러한 문제점을 개선하기 위해 Convolutional Neural Network(CNN) 기반 리뷰 유용성 점수 예측을 통한 새로운 추천 방법론을 제안하였다. 본 연구에서 제안하는 유용한 리뷰를 포함하는 방법론과 기존 모든 선호도 평점을 고려하는 추천 방법론을 비교한 결과, 본 연구에서 제안한 방법론이 더 우수한 예측 성능을 나타내고 있음을 확인할 수 있었다. 또한 본 연구의 결과는 리뷰 유용성에 대한 정보를 개인화 추천 서비스에 반영하면 전통적인 CF의 성능을 향상할 수 있음을 시사한다. Recently, various types of products have been launched with the rapid growth of the e-commerce market. As a result, many users face information overload problems, which is time-consuming in the purchasing decision-making process. Therefore, the importance of a personalized recommendation service that can provide customized products and services to users is emerging. For example, global companies such as Netflix, Amazon, and Google have introduced personalized recommendation services to support users" purchasing decisions. Accordingly, the user"s information search cost can reduce which can positively affect the company"s sales increase. The existing personalized recommendation service research applied Collaborative Filtering (CF) technique predicts user preference mainly use quantified information. However, the recommendation performance may have decreased if only use quantitative information. To improve the problems of such existing studies, many studies using reviews to enhance recommendation performance. However, reviews contain factors that hinder purchasing decisions, such as advertising content, false comments, meaningless or irrelevant content. When providing recommendation service uses a review that includes these factors can lead to decrease recommendation performance. Therefore, we proposed a novel recommendation methodology through CNN-based review usefulness score prediction to improve these problems. The results show that the proposed methodology has better prediction performance than the recommendation method considering all existing preference ratings. In addition, the results suggest that can enhance the performance of traditional CF when the information on review usefulness reflects in the personalized recommendation service.

Mannose Attenuates Colitis-Associated Colorectal Tumorigenesis by Targeting Tumor-Associated Macrophages

Liu Qinglong,Li Xiaojing,Zhang Hao,Li Haitao 대한암예방학회 2022 Journal of cancer prevention Vol.27 No.1

Mannose has recently drawn extensive attention for its substantial anti-cancer activities, but the underlying mechanism remains largely unclear. The aim of this study was to investigate the effects of mannose on experimental colitis-associated colorectal tumorigenesis and underlying mechanisms. Data clearly showed that at plasma concentrations achieved after oral administration, mannose slightly affected malignancy of tumor cells or tumor promoter-induced transformation of pre-neoplastic cells, but substantially suppressed manifestation of the M2-like phenotype of tumor-associated macrophages (TAMs) in a cancer cell and macrophage co-culture model. Mechanistically, mannose might greatly impair the production of tumor cell-derived lactate which has a critical role in the functional polarization of TAMs. Importantly, oral administration of mannose protected mice against colitis-associated colorectal tumorigenesis by normalizing TAM polarization. Collectively, these findings highlight the importance of TAMs in colorectal tumorigenesis, and provide a rationale for introducing mannose supplementation to patients suffering from inflammatory bowel diseases

사용자의 선호도 정보를 활용한 직무 추천 시스템 연구

이청용(Qinglong Li),전상홍(Sanghong Jeon),이창재(Changjae Lee),김재경(Jae Kyeong Kim) 한국IT서비스학회 2021 한국IT서비스학회지 Vol.20 No.3

Recently, online job websites have been activated as unemployment problems have emerged as social problems and demand for job openings has increased. However, while the online job platform market is growing, users have difficulty choosing their jobs. When users apply for a job on online job websites, they check various information such as job contents and recruitment conditions to understand the details of the job. When users choose a job, they focus on various details related to the job rather than simply viewing and supporting the job title. However, existing online job websites usually recommend jobs using only quantitative preference information such as ratings. However, if recommendation services are provided using only quantitative information, the recommendation performance is constantly deteriorating. Therefore, job recommendation services should provide personalized services using various information about the job. This study proposes a recommended methodology that improves recommendation performance by elaborating on qualitative preference information, such as details about the job. To this end, this study performs a topic modeling analysis on the job content of the user profile. Also, we apply LDA techniques to explore topics from job content and extract qualitative preferences. Experiments show that the proposed recommendation methodology has better recommendation performance compared to the traditional recommendation methodology.

A Multi-CNN Model Interacting Contents and Ratings for Predicting Review Helpfulness

Xinzhe Li,Qinglong Li,Jaekyeong Kim 한국지능정보시스템학회 2022 한국지능정보시스템학회 학술대회논문집 Vol.2022 No.6

With the growth of the e-commerce industry, online consumer reviews significantly impact the consumer purchase decision process. Since the consistently increasing number of reviews, the consumer can face an information overload problem. Thus, the consumers have a challenge exploring the information they need. Thus, we argue that predicting the review helpfulness becomes significant. When predicting review helpfulness, since the review contents and star ratings are information written from the same consumer experience, the consistency of the review contents and star ratings is essential. Previous studies predict review helpfulness by considering review content and star ratings simultaneously. However, such an approach has limitations in the representation capacity of star ratings and the capture of the interaction between review content and star ratings. The current study proposed a CNN-CRI mechanism to address the limitations of the previous study. To evaluate the proposed methodology, we utilized real-world online review data from Amazon.com. The results show that our study model indicates better performance than the state-of-the-art approach.

명시적 및 암시적 피드백을 활용한 그래프 컨볼루션 네트워크 기반 추천 시스템 개발

이흠철(Xinzhe Li),김동언(Dongeon Kim),이청용(Qinglong Li),김재경(JaeKyeong Kim) 한국IT서비스학회 2023 한국IT서비스학회지 Vol.22 No.1

With the development of the e-commerce market, various types of products continue to be released. However, customers face an information overload problem in purchasing decision-making. Therefore, personalized recommendations have become an essential service in providing personalized products to customers. Recently, many studies on GCN-based recommender systems have been actively conducted. Such a methodology can address the limitation in disabling to effectively reflect the interaction between customer and product in the embedding process. However, previous studies mainly use implicit feedback data to conduct experiments. Although implicit feedback data improves the data scarcity problem, it cannot represent customers' preferences for specific products. Therefore, this study proposed a novel model combining explicit and implicit feedback to address such a limitation. This study treats the average ratings of customers and products as the features of customers and products and converts them into a high-dimensional feature vector. Then, this study combines ID embedding vectors and feature vectors in the embedding layer to learn the customer-product interaction effectively. To evaluate recommendation performance, this study used the MovieLens dataset to conduct various experiments. Experimental results showed the proposed model outperforms the state-of-the-art. Therefore, the proposed model in this study can provide an enhanced recommendation service for customers to address the information overload problem.

Transcatheter Arterial Chemoembolization Combined with Interferon-α is Safe and Effective for Patients with Hepatocellular Carcinoma after Curative Resection

Zuo, Chaohui,Xia, Man,Liu, Jingshi,Qiu, Xiaoxin,Lei, Xiong,Xu, Ruocai,Liu, Hanchun,Li, Jianliang,Li, Yongguo,Li, Qinglong,Xiao, Hua,Hong, Yuan,Wang, Xiaohong,Zhu, Haizhen,Wu, Qunfeng,Burns, Michael,Li Asian Pacific Journal of Cancer Prevention 2015 Asian Pacific journal of cancer prevention Vol.16 No.1

Objectives: Intrahepatic recurrence is the major cause of death among patients with hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC) after curative surgical resection. Several approaches have been reported to decrease the recurrence rate. The objective of our study was to compare the clinical effects of transcatheter arterial chemoembolization (TACE) combined with interferon-alpha (IFN-${\alpha}$) therapy on recurrence after hepatic resection in patients with HBV-related HCC with that of TACE chemotherapy alone. Methods: We retrospectively analyzed the data from 228 patients who were diagnosed with HBV-related HCC and underwent curative resection between January 2001 to December 2008. The patients were divided into TACE (n = 126) and TACE-IFN-${\alpha}$ (n = 102) groups for postoperative chemotherapy. The TACE regimen consisted of 5-fluorouracil (5-FU), cisplatin (DDP), and the emulsion mixed with mitomycin C (MMC) and lipiodol. The recurrence rates, disease-free survival (DFS), overall survival (OS), and risk of recurrence were evaluated. Results: The clinicopathological parameters and adverse effects were similar between the 2 groups (P > 0.05). The median OS for the TACE-IFN-${\alpha}$ group (36.3 months) was significantly longer than that of the TACE group (24.5 months, P < 0.05). The 3-and 5-year OS for the TACE-IFN-${\alpha}$ group were significantly longer than those of the TACE group (P < 0.05) and the recurrence rate was significantly lower (P < 0.05). The TACE and IFN-${\alpha}$ combination therapy, active hepatitis HBV infection, the number of tumor nodules, microvascular invasion, liver cirrhosis, and the BCLC stage were independent predictors of OS and DFS. Conclusions: The use of the TACE and IFN-${\alpha}$ combination chemotherapy after curative hepatic resection safely and effectively improves OS and decreases recurrence in patients with HBV-related HCC who are at high risk. Our findings can serve as a guide for the selection of postoperative adjuvant chemotherapy for patients with HBV-related HCC who are at high risk of recurrence.

레스토랑 카테고리와 온라인 소비자 리뷰를 이용한 딥러닝 기반 레스토랑 추천 시스템 개발

구하은,LI QINGLONG,김재경 한국경영정보학회 2023 Information systems review Vol.25 No.1

최근에는 외식 산업의 발달과 레스토랑 수요의 증가로 인해 레스토랑 추천 시스템 연구가 활발하게제안되고 있다. 기존 레스토랑 추천 시스템 연구는 정량적인 평점 정보 또는 온라인 리뷰의 감성분석을통해 소비자의 선호도 정보를 추출하였는데 이는 소비자의 의미론적 선호도 정보는 반영하지 못한다는한계가 존재한다. 또한, 레스토랑이 포함하는 세부적인 속성을 반영한 추천 시스템 연구는 부족한실정이다. 이를 해결하기 위해 본 연구에서는 소비자의 선호도와 레스토랑 속성 간의 상호작용을효과적으로 학습할 수 있는 딥러닝 기반 모델을 제안하였다. 먼저, 합성곱 신경망을 온라인 리뷰에적용하여 소비자의 의미론적 선호도 정보를 추출했고, 레스토랑 정보에 임베딩 기법을 적용하여레스토랑의 세부적인 속성을 추출했다. 최종적으로 요소별 연산을 통해 소비자 선호도와 레스토랑속성 간의 상호작용을 학습하여 소비자의 선호도 평점을 예측했다. 본 연구에서 제안한 모델의 추천성능을 평가하기 위해 Yelp.com의 온라인 리뷰를 사용한 실험 결과, 기존 연구의 다양한 모델과 비교했을때 본 연구의 제안 모델이 우수한 추천 성능을 보이는 것을 확인하였다. 본 연구는 레스토랑 산업의빅데이터를 활용한 맞춤형 레스토랑 추천 시스템을 제안함으로써 레스토랑 연구 분야와 온라인 서비스제공자에게 학술적 및 실무적 측면에서 다양한 시사점을 제공할 수 있을 것으로 기대한다. Research on restaurant recommender systems has been proposed due to the development of the food service industry and the increasing demand for restaurants. Existing restaurant recommendation studies extracted consumer preference information through quantitative information or online review sensitivity analysis, but there is a limitation that it cannot reflect consumer semantic preference information. In addition, there is a lack of recommendation research that reflects the detailed attributes of restaurants. To solve this problem, this study proposed a model that can learn the interaction between consumer preferences and restaurant attributes by applying deep learning techniques. First, the convolutional neural network was applied to online reviews to extract semantic preference information from consumers, and embedded techniques were applied to restaurant information to extract detailed attributes of restaurants. Finally, the interaction between consumer preference and restaurant attributes was learned through the element-wise products to predict the consumer preference rating. Experiments using an online review of Yelp.com to evaluate the performance of the proposed model in this study confirmed that the proposed model in this study showed excellent recommendation performance. By proposing a customized restaurant recommendation system using big data from the restaurant industry, this study expects to provide various academic and practical implications.

Rutile TiO₂ nanowire-based perovskite solar cells

Jiang, Qinglong,Sheng, Xia,Li, Yingxuan,Feng, Xinjian,Xu, Tao The Royal Society of Chemistry 2014 Chemical communications Vol.50 No.94

<P>Different lengths of rutile TiO<SUB>2</SUB> nanowires (NW) with wide-open space for effective material filling were used as photoanodes for perovskite solar cells. Cells with 900 nm nanowires as photoanodes exhibit a current density of 22 mA cm<SUP>−2</SUP> and an efficiency of 11.7%, outperforming the reported TiO<SUB>2</SUB> nanowire-based perovskite solar cells.</P> <P>Graphic Abstract</P><P>Different lengths of rutile TiO<SUB>2</SUB> nanowires (NW) with wide-open space for effective material filling were used as photoanodes for perovskite solar cells. <IMG SRC='http://pubs.rsc.org/services/images/RSCpubs.ePlatform.Service.FreeContent.ImageService.svc/ImageService/image/GA?id=c4cc07367c'> </P>