FISM: factored item similarity models for top-N recommender systems

By Santosh Kabbur, Xia Ning, George Karypis. Knowledge Discovery and Data Mining, 2013

732 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Information Systems, Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/0c0ff60e9d39c203929457d1ac3f840f8c8e9619

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Information Systems

In Management Science and Operations Research

Citations in the corpus

A Survey on Accuracy-Oriented Neural Recommendation: From Collaborative Filtering to Information-Rich Recommendation (2021, background): “, Factored Item Similarity Model (FISM) pools the interacted item embeddings as a user representation vector [20], and SVD++ [24] adds UID embedding pu with the interaction history embedding (i.”
Model-Agnostic Counterfactual Reasoning for Eliminating Popularity Bias in Recommender System (2020, background): “rsonalized recommendation has revolutionized a myriad of online applications such as e-commerce [48, 50], search engines [38], and conversational systems [26, 40]. A huge number of recommender models [16, 23, 43] have been developed, for which the default optimization choice is reconstructing histor”
LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation (2020, methodology): “g with neural networks. BeyondmerelyusingIDinformation,anothertypeofCFmethods considers historical items as the pre-existing features of a user, towards better user representations. For example, FISM [21] and SVD++ [25] use the weighted average of the ID embeddings of historical items as the target ”
In2Rec: Influence-based Interpretable Recommendation (2019, methodology): “Inspired by the successes of attention-based deep learning due to its strong ability of representation, researchers presented neural attentive interpretable recommendation model (NAIRS) [30] to extend the traditional item similarity model [13].”
Long and Short-Term Recommendations with Recurrent Neural Networks (2017, methodology): “Fossil[8] is inspired by FPMC and by FISM[11]: instead of factorizing the user-item matrix, it factorizes the item-item matrix.”
How Does NLP Benefit Legal System: A Summary of Legal Artificial Intelligence (2020, methodology): “Many machine learning methods have also been applied for IR like SVD (Xu et al., 2010) or factorization (Rendle, 2010; Kabbur et al., 2013).”
Transfer Learning for Semisupervised Collaborative Recommendation (2016, background): “…due to the dominant research trend of making use of explicit feedback from the very beginning of the collaborative filtering history in the early 1990s [Goldberg et al. 1992] to late 2000s [Koren 2008], and the recent trend of exploiting implicit feedback [Rendle et al. 2009; Kabbur et al. 2013].”
Transfer Learning for Behavior Ranking (2017, methodology): “, we may freely use some off-the-shelf OCCF tools [Rendle et al. 2009; Hu et al. 2008; Kabbur et al. 2013].”
Self-Attentive Sequential Recommendation (2018, methodology): “Following the reduction operations above for FMC, and adding an explicit user embedding (via concatenation), SASRec is equivalent to FPMC. • Factorized Item Similarity Models [8]: FISM estimates a preference score toward item i by considering the similarities between i and items the user consumed be”
Adaptive Bayesian personalized ranking for heterogeneous implicit feedbacks (2015, methodology): “So far, various collaborative filtering algorithms have been proposed, including, (1) maximum margin matrix factorization (MMMF) [30], probabilistic matrix factorization (PMF) [26], SVD++ [11] and Fuzzy-based Telecom Product Recommender System (FTCP-RS) [34] for homogeneous explicit feedbacks, (2) o”
Transfer Learning for Heterogeneous One-Class Collaborative Filtering (2016, methodology): “In FISM [3], in order to estimate the preference of user u on itemi, we learn a similarity between item i and apreferred item i, i.”
Collaborative Denoising Auto-Encoders for Top-N Recommender Systems (2016, methodology): “This processing method is widely used in previous work on recommendation with implicit feedback (e.g., [15, 31, 7]).”
Disentangled Graph Collaborative Filtering (2020, background): “that, later studies like SVD++ [19], FISM [16], NAIS [11], and ACF [3] view personal history of a user as her features, and integrate embeddings of historical items via the average [16, 19] or attention network [3, 11] as user embeddings, showing promise in representation learning.”
Hierarchical Gating Networks for Sequential Recommendation (2019, methodology): “o form the group-level representation. 4.2 Item-item Product The relation between two single items is an important factor to model in the recommendation task and has been widely studied in many years [17, 25], e.g., item-based collaborative filtering methods utilizing the rating vectors of two items”
Adversarial Personalized Ranking for Recommendation (2018, background): “In particular, CDAE and NeuMF are the recently proposed neural recommender models which have shown significant improvements over conventional shallow meth-ods like MF and FISM [19].”
NAIS: Neural Attentive Item Similarity Model for Recommendation (2018, methodology): “Pointwise L2R methods either optimize a regression-based squared loss [1], [35] or classification-based log loss [5], by either sam-”
Fusing Similarity Models with Markov Chains for Sparse Sequential Recommendation (2016, background): “further explored the low-rank property of the similarity matrix to handle sparse datasets [2].”