How to select and weight context dimensions conditions for context-aware recommendation?
Section snippets
Introduction and motivation
Recommender Systems (RS) are specific types of information filtering to present items (movies, music, books, news, images, etc.) that likely to interest the user (Díaz et al., 2012, Chen et al., 2018, Usha Yadav and Bhatia, 2020). Typically, an RS compares the current situation of a user to some baseline characteristics (his/her history versus those of other users), and tries to predict the rating that a user would provide.
In this respect, context-awareness can solve many problems that affect
Formal definitions
We start this section by defining and modeling some notions like context, user context, etc.
- •
Context dimension: In order to use a context itself, the definition of its associated dimensions must be understood beforehand (Crane, Benachour, & Coulton, 2011). The dimension term is a measurable extent or quantity (such as amplitude, brightness, hue, frequency, length, width, height, time, mass, volume, weight) that denotes the degree to, or range over, which something extends. In our case, a
Related work
We basically built traditional RSs upon the knowledge of user preferences for a particular set of items, as well as the input data for traditional RSs. This input is typically based on the records of the form . In contrast, CARS are built based on the knowledge of context user preferences and typically deal with data records of the form .
An RS relies on a content-based approach or Collaborative Filtering (CF) approach or even both. The operating
SWCC: A new prediction model based on the selection and the weighting of context conditions
From the related work, we notice that context condition selection and weighting are important for improving prediction in CARS. Nevertheless, they are used separately. In this paper, we introduce a prediction model (c.f, Fig. 1) that starts from a current situation and a user history to predict the rating of the current item. The first step in our approach comprises selecting the context conditions. We propose four selection methods (cf. sub-section 4.1.1). We base the first one on the variance
Experimental study
In this section, we usher by describing used datasets, and we specify how these datasets were split in terms of training and test. Then, we present and discuss the obtained results.
Conclusions and future work
Many studies have investigated contextual conditions weighting and contextual conditions selection independently, although these two topics seem closely related. In this paper, we introduce an algorithm that combines contextual conditions selection and weighting in context-aware recommendations. In fact, we address the problem of how to select contextual conditions in order to eliminate the irrelevant ones, how to associate importance degrees to a contextual conditions context and how to
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgement
Sadok Ben Yahia is supported by the Astra funding program Grant 2014-2020.4.01.16-032.
References (96)
- et al.
Situational computing: An innovative architecture with imprecise reasoning
Journal of Systems and Software
(2007) - et al.
A review of methods for capacity identification in choquet integral based multi-attribute utility theory: Applications of the kappalab r package
European Journal of Operational Research
(2008) - et al.
Rudiments of rough sets
Information Science
(2007) A machine learning based robust prediction model for real-life mobile phone data
Internet Things
(2019)- Abdelwahab, A., Matsuba, I., Horiuchi, Y., & Kuroiwa, S. (2012). Feature optimization approach for improving the...
- et al.
Towards a better understanding of context and context-awareness
- Adomavicius, G., Sankaranarayanan, R., Sen, S., & Tuzhilin, A. (2005). Incorporating contextual information in...
- et al.
Towards more confident recommendations: Improving recommender systems using filtering approach based on rating variance
- et al.
Multi-criteria recommender systems
- et al.
Toward the next generation of recommender systems: A survey of the state-of-the-art and possible extensions
IEEE Transactions on Knowledge and Data Engineering
(2005)
Context-aware recommender system based on boolean matrix factorisation
A collaborative filtering recommender system using genetic algorithm
Information Processing & Management
Incarmusic: Context-aware music recommendations in a car
Context relevance assessment and exploitation in mobile recommender systems
Personal and Ubiquitous Computing
Item weighting techniques for collaborative filtering
A generic framework for comparing semantic similarities on a subsumption hierarchy
Following the user’s interests in mobile context-aware recommender systems:the hybrid-∊-greedy algorithm
Selective contextual information acquisition in travel recommender systems
Information Technology & Tourism
Empirical analysis of predictive algorithms for collaborative filtering
Business context information manager: An approach to improve information systems
Personalized e-tourism attraction recommendation based on context
A survey of collaborative filtering-based recommender systems: From traditional methods to hybrid methods based on social networks
IEEE Access
Conrec: A software framework for context-aware recommendation based on dynamic and personalized context
A user study of the spatial and temporal dimensions of context to support virtual learning environments
Analytical games for knowledge engineering of expert systems in support to situational awareness: The reliability game case study
Expert Systems With Applications
Mobile and context-aware geobi applications: A multilevel model for structuring and sharing of contextual information
Journal of Geographic Information System
Using offline metrics and user behavior analysis to combine multiple systems for music recommendation
Bayesian network classifiers
Machine Learning
Contextual information based recommender system using singular value decomposition
The movielens datasets: History and context
An algorithmic framework for performing collaborative filtering
General factorization framework for context-aware recommendations
CoRR
A popularity-based neighbor selection model in p2p file-sharing system
Context-aware recommendation using rough set model and collaborative filtering
Artificial Intelligence Review
Semantic similarity based on corpus statistics and lexical taxonomy
An automatic weighting scheme for collaborative filtering
Particle filter with swarm move for optimization
Two algorithms for approximate string matching in static texts
Asynchronous distributed matrix factorization with similar user and item based regularization
Multiverse recommendation: n-dimensional tensor factorization for context-aware collaborative filtering
Cited by (5)
Design Science Research in Developing Traffic Urgency Model from Text for Determining Transportation Complaint Priority - An Initial Investigation
2023, 2023 International Seminar on Intelligent Technology and Its Applications: Leveraging Intelligent Systems to Achieve Sustainable Development Goals, ISITIA 2023 - ProceedingAnalyzing Traffic-Complaint Tweets Based on Time-Location Context Analysis to Develop A Traffic Urgency Model
2023, 2023 8th International Conference on Business and Industrial Research, ICBIR 2023 - ProceedingsContext-Aware Recommendation System Survey: Recommendation When Adding Contextual Information
2022, Proceeding - 6th International Conference on Information Technology, Information Systems and Electrical Engineering: Applying Data Sciences and Artificial Intelligence Technologies for Environmental Sustainability, ICITISEE 2022Contextual and Sentimental Teachers' Peer Recommendations
2022, 2022 IEEE 25th International Conference on Computer Supported Cooperative Work in Design, CSCWD 2022