[Paper] Personalized Recommendation of Tumblr Posts Using Graph Convolutional Networks with Preference-aware Multimodal Features

Ohtomo, Kazuma; Harakawa, Ryosuke; Ogawa, Takahiro; Haseyama, Miki; Iwahashi, Masahiro

doi:10.3169/mta.9.54

Cited by 7 publications

(6 citation statements)

References 31 publications

(50 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on the personalized recommendation technology, a personalized music recognition system is constructed for the user's personal data, music preference, emotional characteristics, and other data [6]. Ohtomo et al use based on preference aware multimodal feature graph convolution network and verified the effectiveness of this method by using real data sets [7]. Li et al designed an intelligent learning system through the content recommendation and realized the personalized recommendation of network teaching resources [8].…”

Section: Related Workmentioning

confidence: 99%

“…In formula (7), x and y represent the scores of users u and v on project resources; r ui and r vi represent the resource i, and n represents the number of project resources. After obtaining the average value of all category data points, the data point closest to the average value point can be used as the new clustering center, as shown in the following formula.…”

Section: E X Ymentioning

confidence: 99%

See 1 more Smart Citation

Intelligent Sports Auxiliary Training Method Based on Collaborative Filtering Recommendation Algorithm

Sun

She

2022

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

In recent years, people’s living standards are rising, and the demand for health is also rising. National physical exercise has become the current trend. Using personalized recommendation technology to screen appropriate information and assist people in sports training can improve the efficiency of users to obtain relevant information and improve users’ physical quality. Therefore, the collaborative filtering recommendation algorithm is optimized by combining bisection, and the intelligent recommendation model of sports training resources is constructed based on the algorithm. This paper uses the improved algorithm to calculate the similarity between users. Compared with other traditional algorithms, the user algorithm in this paper has higher accuracy and certain reference value. To sum up, the intelligent recommendation model of sports training resources proposed in the research can more accurately recommend suitable sports training resources for users, so as to assist users in correct sports training and improve users’ physical quality, which plays an improvement of the physical quality of the whole people in China.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: E X Ymentioning

confidence: 99%

Intelligent Sports Auxiliary Training Method Based on Collaborative Filtering Recommendation Algorithm

Sun

She

2022

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

show abstract

“…Chaabi's research group proposed a user interest model from a generalized to a specific hierarchy, which can effectively differentiate the interest characteristics of different classes [17]. Ohtomo et al argue that users' interests when reading news can be divided into two categories, short term interests and long-term interests, with short term interests tending to be related to the timeliness of popular information and changing rapidly, while longer term interests tend to reflect users' real interests [18]. In contrast, Chen's team used a multi-strategy machine to build long-term and short-term interest patterns, and through an in-depth analysis of this system, a new content-based fusion algorithm was proposed and designed [19].…”

Section: Related Workmentioning

confidence: 99%

Design of an English Web-based Teaching Resource Sharing Platform based on Mobile Web Technology

Zhang¹

2023

IJACSA

View full text Add to dashboard Cite

Thanks to the booming technology of computers and multimedia, student-centered online teaching resource platforms have become an important way for students to learn. However, English teaching resource platforms at the present stage fail to effectively integrate the massive and scattered learning resources. Based on this, the study proposes an English online teaching resource sharing platform based on mobile Web technology, using the SOAP protocol to deploy heterogeneous data resources as Web services to achieve interchangeability between heterogeneous resources. In addition, to enhance the efficient use of learning resources by students, the study proposes a hybrid algorithm based on collaborative filtering algorithm and sequential pattern mining algorithm to achieve personalized sequential recommendation for students. The results show that the platform created by the study exhibits excellent performance in terms of resource transfer capability, achieves efficient teaching resource sharing in a short response time and also shows that the proposed recommendation algorithm is highly accurate.

show abstract

“…We note that developing a sequential recommendation that considers the irrelevant modality remains challenging and has hardly been investigated in previous studies. Existing methods [18]- [27] cannot estimate the irrelevant modality because they perform the early fusion of multi-modal features through, for example, linear transformation and concatenation. The recommendation performance is promising if the multi-modal features are orderly; i.e., there is no irrelevant modality.…”

Section: Introductionmentioning

confidence: 99%

AM-Bi-LSTM: Adaptive Multi-Modal Bi-LSTM for Sequential Recommendation

Ohtomo,

Harakawa,

Iisaka

et al. 2024

IEEE Access

Self Cite

View full text Add to dashboard Cite

Conventional methods for the early fusion of multi-modal features cannot recognize the relevant modality corresponding to the demand of each user in sequential recommendation. In this paper, we propose the adaptive multi-modal bidirectional long short-term memory network (AM-Bi-LSTM) to recognize the relevant modality for sequential recommendation. Specifically, we construct a new recurrent neural network model that is based on the bidirectional long short-term memory network and obtains multimodal features, including each user's sequential actions. Our new modality attention module calculates the importance degree of multi-modal features for sequential operations via the late-fusion approach, which results in the method recognizing the relevant modality. In experiments on a multi-modal and sequential dataset including 14,941 clicks constructed from the largest Web service for teachers in Japan, we demonstrate that AM-Bi-LSTM outperforms existing methods in terms of the diversity, explainability, and accuracy of recommendation. Specifically, we obtain Recall@10 that is 0.1005 better than that of existing early-fusion methods. Moreover, we obtain a value of catalog coverage@10 (representing diversity) that is 0.1710 higher than that for existing methods.

show abstract

[Paper] Personalized Recommendation of Tumblr Posts Using Graph Convolutional Networks with Preference-aware Multimodal Features

Cited by 7 publications

References 31 publications

Intelligent Sports Auxiliary Training Method Based on Collaborative Filtering Recommendation Algorithm

Intelligent Sports Auxiliary Training Method Based on Collaborative Filtering Recommendation Algorithm

Design of an English Web-based Teaching Resource Sharing Platform based on Mobile Web Technology

AM-Bi-LSTM: Adaptive Multi-Modal Bi-LSTM for Sequential Recommendation

Contact Info

Product

Resources

About