A microblog recommendation algorithm based on social tagging and a temporal interest evolution model

Sun, Xiaoyan; Li, Xing-xing; Xu, Dongrong; Yuan, Zhigang; Huang, Chi En

doi:10.1631/fitee.1400368

Cited by 14 publications

(4 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The result indicates that temporal patterns including the order and the timing of user enrollment behaviors help to course recommendation. A possible explanation is that the order and the timing of user enrollment behaviors capture two important kinds of learning interests: interest drift and interest evolution [12][13][14]. On the one hand, users often enroll in many courses in a short period of time, which usually implies users' interest drift.…”

Section: The Role Of Time-aware Positional Encodingmentioning

confidence: 99%

“…For example, one may be interested in courses related to blockchain given its popularity, while one's interest may turn to epidemiology during the COVID-19 pandemic. Interest evolution indicates users' evolving learning interests as they gain knowledge, skills, and experience over time [13,14]. For example, one is likely to study "C++ Advanced Programming" after studying the course of "C++ Basic Programming" since the former is an advanced course of the latter.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Profiling temporal learning interests with time-aware transformers and knowledge graph for online course recommendation

Zhou

Jiang

et al. 2022

Electron Commer Res

View full text Add to dashboard Cite

Profiling users' temporal learning interests is key to online course recommendation. Previous studies mainly profile users' learning interests by aggregating their historical behaviors with simple fusing strategies, which fails to capture their temporal interest patterns underlying the sequential user behaviors. To fill the gap, we devise a recommender that incorporates time-aware Transformers and a knowledge graph to better capture users' temporal learning interests. First, we introduce stacked Transformers to extract users' temporal learning interests underlying users' course enrollment sequences. In addition, we design a time-aware positional encoding module to consider the enrollment time intervals between courses. Third, we incorporate a knowledge graph to utilize the latent knowledge connections between courses. The proposed method outperforms state-of-the-art baselines for course recommendation. Furthermore, findings in the ablation study offers several insights for future research. The proposed model can be implemented in online learning platforms to increase user engagement and reduce dropout rate.

show abstract

Section: The Role Of Time-aware Positional Encodingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Profiling temporal learning interests with time-aware transformers and knowledge graph for online course recommendation

Zhou

Jiang

et al. 2022

Electron Commer Res

View full text Add to dashboard Cite

show abstract

“…Recommendation algorithms [4] are another common method to predict the purchase intention of ECP users. Based on various social tags, Yuan et al [5] designed a collaborative filtering (CF) recommendation algorithm for Sina Weibo users. Nanopoulos et al [6] classified songs by social tags, and proved that the classification based on social tags is more accurate than that based on texts and other data sources.…”

Section: Recommendation Algorithmsmentioning

confidence: 99%

Prediction of Purchase Intention among E-Commerce Platform Users Based on Big Data Analysis

Guo¹,

Yang²,

Tian³

2020

RIA

View full text Add to dashboard Cite

The boom of e-commerce platforms (ECPs) has created a massive amount of data on user behaviors. To realize precision marketing, the ECPs must mine out the effective information from the massive data, and predict the purchase intention of their users. Therefore, this paper attempts to design an effective prediction model of purchase intention among ECP users. Firstly, feature engineering, coupled with big data analysis, was performed to identify the features that directly bear on the purchase intention of ECP users. Drawing on these features, two prediction models were established based on linear regression (LR) and extreme gradient boosting (XGBoost), respectively. The XGBoost model was found to be more effective through experiment on ECP users using cellphones. Finally, the prediction effects of the XGBoost-based prediction model were verified through an experiment on Epinions Trust Network Dataset. The research results provide new insights into user behaviors on ECPs.

show abstract

“…Along with the rapidly development of cloud computing and big data, overload of data and information is increasingly obvious, which results in lower accuracy and efficiency of relevant contents and items search. Thus, RSs are of vital importance in human being's daily life [24]. Generally, RSs are used to predict users' affection for items and then recommend the most appropriate items (Top-N items) to users according to their affections.…”

Section: Related Workmentioning

confidence: 99%

NE-UserCF: Collaborative Filtering Recommender System Model based on NMF and E2LSH

Yun¹

2017

IJPE

View full text Add to dashboard Cite

With the rapid development of big data and cloud computing, recommender systems (RSs) have gained significant attention in recent decades. However, there are still many challenges and drawbacks existed in RSs, such as complex and high-dimensional data, low recommendation accuracy, time-consuming and low-efficiency, which to a large extent restrict its applications. Non-negative Matrix Factorization algorithm (NMF) is a matrix factorization algorithm which finds the positive factorization of a given positive matrix. It can eliminate invalid and redundant features in user-rating matrix (URM), reduce URM's dimension. Exact Euclidean Locality Sensitive Hashing (E 2 LSH) is an advanced algorithm for solving the approximate or exact Near Neighbor Search in high dimensional spaces. It can cluster similar-interest users (SIUs) of URM efficiently. Therefore, the authors propose an improved recommender system model named NE-UserCF (NMF-E 2 LSH-UserCF) based on NMF and E 2 LSH to improve the quality and performance of recommendation. The authors first utilize the NMF to process original URM, get a new-URM without invalid and redundant features. Then use E 2 LSH to cluster users in new-URM based on their interests and produce the similar-interest-user matrix (SIUM). The authors further process the Top-10 recommendations by adopting the user-based collaborative filtering algorithm (UserCF). Finally evaluate experimental results by analyzing metrics Precision, Recall, Coverage and Popularity. Experiments indicate that NE-UserCF proposed in this paper improves the quality of recommendation and has a good performance.

show abstract

A microblog recommendation algorithm based on social tagging and a temporal interest evolution model

Cited by 14 publications

References 29 publications

Profiling temporal learning interests with time-aware transformers and knowledge graph for online course recommendation

Profiling temporal learning interests with time-aware transformers and knowledge graph for online course recommendation

Prediction of Purchase Intention among E-Commerce Platform Users Based on Big Data Analysis

NE-UserCF: Collaborative Filtering Recommender System Model based on NMF and E2LSH

Contact Info

Product

Resources

About