Modeling the Heterogeneous Duration of User Interest in Time-Dependent Recommendation: A Hidden Semi-Markov Approach

Zhang, Haidong; Ni, Wancheng; Li, Xin; Yang, Yiping

doi:10.1109/tsmc.2016.2599705

Cited by 35 publications

(13 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In conformity with the Markov property, every hidden state is conditionally dependent only on the previous one [34]. There already exist a few applications of HMMs to time-aware collaborative filtering [96,119]. In [47], the hidden states is used to model the (unobserved) context of the user.…”

Section: Sequence Modelingmentioning

confidence: 99%

Sequence-Aware Recommender Systems

Quadrana¹,

2018

View full text Add to dashboard Cite

Recommender systems are one of the most successful applications of data mining and machine learning technology in practice. Academic research in the field is historically often based on the matrix completion problem formulation, where for each user-item-pair only one interaction (e.g., a rating) is considered. In many application domains, however, multiple user-item interactions of different types can be recorded over time. And, a number of recent works have shown that this information can be used to build richer individual user models and to discover additional behavioral patterns that can be leveraged in the recommendation process. In this work we review existing works that consider information from such sequentially-ordered useritem interaction logs in the recommendation process. Based on this review, we propose a categorization of the corresponding recommendation tasks and goals, summarize existing algorithmic solutions, discuss methodological approaches when benchmarking what we call sequence-aware recommender systems, and outline open challenges in the area.

show abstract

Section: Sequence Modelingmentioning

confidence: 99%

Sequence-Aware Recommender Systems

Quadrana¹,

2018

View full text Add to dashboard Cite

show abstract

“…In [20,24,25], the hidden (semi-)Markov models are applied for recommender systems, in which the hidden variables are used to describe the states of users and the item selections are the observed evidences. It is a convenient way to use HMMs to describe the item selections of users.…”

Section: Related Workmentioning

confidence: 99%

“…There are probabilistic models such as hidden Markov models (HMMs) [20][21][22][23][24][25][26][27] and Kalman filter [28][29][30][31][32] in recommender systems. These studies explain the behaviors of users in probabilistic models for making prediction.…”

Section: Introductionmentioning

confidence: 99%

Recommendation over time: a probabilistic model of time-aware recommender systems

Lin

Chen

2019

Sci. China Inf. Sci.

View full text Add to dashboard Cite

In time-aware recommender systems, we have to consider the dynamic aspect of recommendation that is fond of new coming data. Usually, the recent data is more closely related to current recommendation tasks and the early data are useful to indicate overall measurements of the preferences. We propose a probabilistic model that uses the early data to generate the prior distribution and the recent data to capture the change of the states of both users and items in collaborative filtering systems. Our model is dynamic in the sense that it updates every time receiving new data. The time cost of every updating has a constant limit, which is suitable to deal with large scale data for online recommendation. Experiments on real datasets show the improvement performance of our model over the existing time-aware recommender systems.

show abstract

“…The proof is thus completed. Notice that there exist coupling terms involved by the Lyapunov matrices and system matrices in condition (22), which bring some obstacle for filter synthesis problem. In [36], a traditional decoupling inequality −P i Z −1 ν P i ≤ Z ν − 2P i , ν = 1, 2 is employed to eliminate the product between the system matrices and Lyapunov variables.…”

Section: B New Bounded Real Lemmamentioning

confidence: 99%

A Novel Memory Filtering Design for Semi-Markovian Jump Time-Delay Systems

Wei

Qiu

Karimi

et al. 2018

IEEE Trans. Syst. Man Cybern, Syst.

100

View full text Add to dashboard Cite

This paper investigates the problem of delay-dependent H∞ memory filtering for continuous-time semi-Markovian jump linear systems (MJLSs) with time-varying delay in an input-output framework. Differing from the constant transition rates (TRs) in the conventional MJLSs, the TRs of the semi-MJLSs depend on the random sojourn-time and are thus with time-varying characteristics. By utilizing a two-term approximation for the terms with time-varying delay, it is first shown that the filtering error system (FES) can be reformulated into a feedback interconnection form and the stability and performance analysis problem of the FES can be recast as the scaled small gain (SSG) problem of an interconnected system. Then, based on a semi-Markovian Lyapunov-Krasovskii formulation of SSG condition combined with projection lemma, the H∞ filter synthesis for the underlying semi-MJLSs is formulated in terms of linear matrix inequalities. Finally, simulation studies are provided to evaluate the effectiveness and superiority of the proposed design method

show abstract

Modeling the Heterogeneous Duration of User Interest in Time-Dependent Recommendation: A Hidden Semi-Markov Approach

Cited by 35 publications

References 47 publications

Sequence-Aware Recommender Systems

Sequence-Aware Recommender Systems

Recommendation over time: a probabilistic model of time-aware recommender systems

A Novel Memory Filtering Design for Semi-Markovian Jump Time-Delay Systems

Contact Info

Product

Resources

About