• Journal of Information Processing Systems
• /
• 제17권4호
• /
• pp.707-720
• /
• 2021

To improve the accuracy of the recommendation system, multi-criteria recommendation systems have been widely researched. However, it is highly complicated to extract the preferred features of users and items from the data. To this end, subjective indicators, which indicate a user's priorities for personalized recommendations, should be derived. In this study, we propose a method for generating recommendation candidates by predicting multi-criteria ratings from reviews and using them to derive user priorities. Using a deep learning model based on convolutional neural network (CNN) and bidirectional long short-term memory (BiLSTM), multi-criteria prediction ratings were derived from reviews. These ratings were then aggregated to form a linear regression model to predict the overall rating. This model not only predicts the overall rating but also uses the training weights from the layers of the model as the user's priority. Based on this, a new score matrix for recommendation is derived by calculating the similarity between the user and the item according to the criteria, and an item suitable for the user is proposed. The experiment was conducted by collecting the actual "TripAdvisor" dataset. For performance evaluation, the proposed method was compared with a general recommendation system based on singular value decomposition. The results of the experiments demonstrate the high performance of the proposed method.

• Journal of Information Processing Systems
• /
• 제17권2호
• /
• pp.369-384
• /
• 2021

In a personalized product recommendation system, when the amount of log data is large or sparse, the accuracy of model recommendation will be greatly affected. To solve this problem, a personalized product recommendation method using deep factorization machine (DeepFM) to analyze user behavior is proposed. Firstly, the K-means clustering algorithm is used to cluster the original log data from the perspective of similarity to reduce the data dimension. Then, through the DeepFM parameter sharing strategy, the relationship between low- and high-order feature combinations is learned from log data, and the click rate prediction model is constructed. Finally, based on the predicted click-through rate, products are recommended to users in sequence and fed back. The area under the curve (AUC) and Logloss of the proposed method are 0.8834 and 0.0253, respectively, on the Criteo dataset, and 0.7836 and 0.0348 on the KDD2012 Cup dataset, respectively. Compared with other newer recommendation methods, the proposed method can achieve better recommendation effect.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권10호
• /
• pp.2809-2821
• /
• 2023

Effective recommendation of similar business groups is a critical factor in obtaining market information for companies. In this study, we propose a novel method for enhancing similar business group recommendation by incorporating derivative criteria and web crawling. We use employment announcements, employment incentives, and corporate vocational training information to derive additional criteria for similar business group selection. Web crawling is employed to collect data related to the derived criteria from 'credit jobs' and 'worknet' sites. We compare the efficiency of different datasets and machine learning methods, including XGBoost, LGBM, Adaboost, Linear Regression, K-NN, and SVM. The proposed model extracts derivatives that reflect the financial and scale characteristics of the company, which are then incorporated into a new set of recommendation criteria. Similar business groups are selected using a Euclidean distance-based model. Our experimental results show that the proposed method improves the accuracy of similar business group recommendation. Overall, this study demonstrates the potential of incorporating derivative criteria and web crawling to enhance similar business group recommendation and obtain market information more efficiently.

• Journal of Information Processing Systems
• /
• 제15권5호
• /
• pp.1141-1155
• /
• 2019

Today, most approaches used in the recommendation system provide correct data prediction similar to the data that users need. The method that researchers are paying attention and apply as a model in the recommendation system is the communities' detection in the big social network. The outputted result of this approach is effective in improving the exactness. Therefore, in this paper, the personalized movie recommendation system that combines data mining for the k-clique method is proposed as the best exactness data to the users. The proposed approach was compared with the existing approaches like k-clique, collaborative filtering, and collaborative filtering using k-nearest neighbor. The outputted result guarantees that the proposed method gives significant exactness data compared to the existing approach. In the experiment, the MovieLens data were used as practice and test data.

• 한국컴퓨터정보학회논문지
• /
• 제28권9호
• /
• pp.93-104
• /
• 2023

현대 사회의 서비스 방식은 대면 방식보다 비대면 방식을 선호하는 추세이다. 하지만 신발과 같이 상품을 추천하는 서비스는 대면 방식의 서비스가 불가피하다. 본 논문에서는 비대면 서비스를 목적으로 자동으로 발의 사이즈를 측정하고, 측정 결과를 기반으로 신발을 추천하는 시스템을 제안한다. 제안방법의 성능을 분석하기 위해 사이즈 측정 오차율과 추천성능을 분석하였다. 추천성능 실험에 사용한 방법은 총 10가지이고, 이의 방법 중 가장 좋은 성능을 보이는 추천 방법을 시스템에 적용하였다. 오차율에 대한 실험결과, 사이즈 관련 오차가 작음을 알 수 있었고, 추천성능에 대한 실험결과, 추천에 대한 유의한 결과를 도출할 수 있었다. 본 논문에서의 제안방법은 실험실 수준으로 향후 실제 환경으로 확대 적용할 필요가 있다.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제10권10호
• /
• pp.399-406
• /
• 2021

콘텐츠 산업의 꾸준한 성장에 따라 수많은 콘텐츠 중에서 개인의 취향에 적합한 콘텐츠를 자동으로 추천하는 연구의 필요성이 증가하고 있다. 콘텐츠 자동 추천의 정확도를 향상시키기 위해서는 콘텐츠에 대한 사용자의 선호 이력을 바탕으로 하는 기존 추천 기법과 더불어 콘텐츠의 메타데이터 및 콘텐츠 자체에서 추출할 수 있는 특징을 융합한 추천 기법이 필요하다. 본 연구에서는 음악의 소리 데이터로부터 태그 정보를 분류하는 LSTM 기반의 모델을 학습하고 분류된 태그 정보를 음악의 메타 데이터로 추가하여, 그래프 임베딩 시 콘텐츠의 특징까지 고려할 수 있는 KPRN 기반의 새로운 콘텐츠 추천 방법을 제안한다. 카카오 아레나 데이터 기반 실험 결과, 본 연구의 제안 방법은 기존의 임베딩 기반 추천 방법보다 우수한 추천 정확도를 보였다.

• 인터넷정보학회논문지
• /
• 제25권1호
• /
• pp.49-56
• /
• 2024

There have been emerging many use-cases applying recommendation systems especially in online platform. Although the performance of recommendation systems is affected by a variety of factors, selecting appropriate features is difficult since most of recommendation systems have sparse data. Conventional matrix factorization (MF) method is a basic way to handle with problems in the recommendation systems. However, the MF based scheme cannot reflect non-linearity characteristics well. As deep learning technology has been attracted widely, a deep neural network (DNN) framework based collaborative filtering (CF) was introduced to complement the non-linearity issue. However, there is still a problem related to feature embedding for use as input to the DNN. In this paper, we propose an effective method using singular value decomposition (SVD) based feature embedding for improving the DNN performance of recommendation algorithms. We evaluate the performance of recommendation systems using MovieLens dataset and show the proposed scheme outperforms the existing methods. Moreover, we analyze the performance according to the number of latent features in the proposed algorithm. We expect that the proposed scheme can be applied to the generalized recommendation systems.

• 한국ITS학회 논문지
• /
• 제8권1호
• /
• pp.119-126
• /
• 2009

최근 개개인의 취향이나 특성을 고려하여 자동으로 사용자에게 정보를 찾아주거나 추천해주는 추천 서비스 시스템이 많이 개발되고 있다. 하지만 사용자의 상황에 따른 선호도를 고려하지 않을 경우 정확한 추천이 힘든 단점이 있다. 따라서 본 논문에서는 사용자의 상황에 따른 선호도를 고려하여 정확한 추천을 할 수 있는 필터링 방법을 제안하였다. 이를 위해 상황에 따른 사용자 선호도를 구하고 피어슨 상관계수를 이용하여 사용자의 상황별 오브젝트 선호도를 구하였다. 실험 결과, 기존의 서비스 시스템들과 비교하여 precision은 11%, 2%, recall은 8%, 4% 향상되었으며, 전체적으로 precision은 77%, recall은 53%로 나타났다.

• 한국HCI학회:학술대회논문집
• /
• 한국HCI학회 2007년도 학술대회 1부
• /
• pp.268-273
• /
• 2007

In this paper, we present effective methods for automatic music recommendation system which automatically recommend music by signal processing technology. Conventional music recommendation system use users’ music downloading pattern, but the method does not consider acoustic characteristics of music. Sometimes, similarities between music are used to find similar music for recommendation in some method. However, the feature used for calculating similarities is not highly related to music characteristics at the system. Thus, our proposed method use high-level music characteristics such as rhythm pattern, timbre characteristics, and the lyrics. In addition, our proposed method store features of music, which individuals queried, to recommend music based on individual taste. Experiments show the proposed method find similar music more effectively than a conventional method. The experimental results also show that the proposed method could be used for real-time application since the processing time for calculating similarities between music, and recommending music are fast enough to be applicable for commercial purpose.

• Journal of Information Processing Systems
• /
• 제15권3호
• /
• pp.616-631
• /
• 2019

With the rapid increase of information on the World Wide Web, finding useful information on the internet has become a major problem. The recommendation system helps users make decisions in complex data areas where the amount of data available is large. There are many methods that have been proposed in the recommender system. Collaborative filtering is a popular method widely used in the recommendation system. However, collaborative filtering methods still have some problems, namely cold-start problem. In this paper, we propose a movie recommendation system by using social network analysis and collaborative filtering to solve this problem associated with collaborative filtering methods. We applied personal propensity of users such as age, gender, and occupation to make relationship matrix between users, and the relationship matrix is applied to cluster user by using community detection based on edge betweenness centrality. Then the recommended system will suggest movies which were previously interested by users in the group to new users. We show shown that the proposed method is a very efficient method using mean absolute error.