Hybrid Filtering Recommendation System in an Educational Context

In education, the needs of learners are different in the majority of the time, as each has specificities in terms of preferences, performance and goals. Recommendation systems have proven to be an effective way to ensure this learning personalization. Already used and tested in other areas such as e-commerce, their adaptation to the educational context has led to several research studies that have tried to find the best approaches with the best expected results. This article suggests that a hybridization of recommendation systems filtering methods can improve the quality of recommendations. An experiment was conducted to test an approach that combines content-based filtering and collaborative filtering. The results proved to be convincing.

Exploiting Latent Information in Recommender Systems

<p>This thesis exploits latent information in personalised recommendation, and investigates how this information can be used to improve recommender systems. The investigations span three directions: scalar rating-based collaborative filtering, distributional rating-based collaborative filtering, and distributional ratingbased hybrid filtering. In the first investigation, the thesis discovers through data analysis three problems in nearest neighbour collaborative filtering — item irrelevance, preference imbalance, and biased average — and identifies a solution: incorporating “target awareness” in the computation of user similarity and rating deviation. Two new algorithms are subsequently proposed. Quantitative experiments show that the new algorithms, especially the first one, are able to significantly improve the performance under normal situations. They do not however excel in cold-start situations due to greater demand of data. The second investigation builds upon the experimental analysis of the first investigation, and examines the use of discrete probabilistic distributional modelling throughout the recommendation process. It encompasses four ideas: 1) distributional input rating, which enables the explicit representation of noise patterns in user inputs; 2) distributional voting profile, which enables the preservation of not only shift but also spread and peaks in user’s rating habits; 3) distributional similarity, which enables the untangled and separated similarity computation of the likes and the dislikes; and 4) distributional prediction, which enables the communication of the uncertainty, granularity, and ambivalence in the recommendation results. Quantitative experiments show that this model is able to improve the effectiveness of recommendation compared to the scalar model and other published discrete probabilistic models, especially in terms of binary and list recommendation accuracy. The third investigation is based on an analysis regarding the relationship between rating, item content, item quality, and “intangibles”, and is enabled by the discrete probabilistic model proposed in the second investigation. Based on the analysis, a fundamentally different hybrid filtering structure is proposed, where the hybridisation strategy is neither linear nor sequential, but of a divide-and-conquer shape backed by probabilistic derivation. Experimental results show that it is able to outperform the standard linear and sequential hybridisation structures.</p>

Exploiting Latent Information in Recommender Systems

<p>This thesis exploits latent information in personalised recommendation, and investigates how this information can be used to improve recommender systems. The investigations span three directions: scalar rating-based collaborative filtering, distributional rating-based collaborative filtering, and distributional ratingbased hybrid filtering. In the first investigation, the thesis discovers through data analysis three problems in nearest neighbour collaborative filtering — item irrelevance, preference imbalance, and biased average — and identifies a solution: incorporating “target awareness” in the computation of user similarity and rating deviation. Two new algorithms are subsequently proposed. Quantitative experiments show that the new algorithms, especially the first one, are able to significantly improve the performance under normal situations. They do not however excel in cold-start situations due to greater demand of data. The second investigation builds upon the experimental analysis of the first investigation, and examines the use of discrete probabilistic distributional modelling throughout the recommendation process. It encompasses four ideas: 1) distributional input rating, which enables the explicit representation of noise patterns in user inputs; 2) distributional voting profile, which enables the preservation of not only shift but also spread and peaks in user’s rating habits; 3) distributional similarity, which enables the untangled and separated similarity computation of the likes and the dislikes; and 4) distributional prediction, which enables the communication of the uncertainty, granularity, and ambivalence in the recommendation results. Quantitative experiments show that this model is able to improve the effectiveness of recommendation compared to the scalar model and other published discrete probabilistic models, especially in terms of binary and list recommendation accuracy. The third investigation is based on an analysis regarding the relationship between rating, item content, item quality, and “intangibles”, and is enabled by the discrete probabilistic model proposed in the second investigation. Based on the analysis, a fundamentally different hybrid filtering structure is proposed, where the hybridisation strategy is neither linear nor sequential, but of a divide-and-conquer shape backed by probabilistic derivation. Experimental results show that it is able to outperform the standard linear and sequential hybridisation structures.</p>

The Hybrid Recommender System of the Indonesian Online Market Products using IMDb weight rating and TF-IDF

Today, consumers are faced with an abundance of information on the internet; accordingly, it is hard for them to reach the vital information they need. One of the reasonable solutions in modern society is implementing information filtering. Some researchers implemented a recommender system as filtering to increase customers’ experience in social media and e-commerce. This research focuses on the combination of two methods in the recommender system, that is, demographic and content-based filtering, commonly it is called hybrid filtering. In this research, item products are collected using the data crawling method from the big three e-commerce in Indonesia (Shopee, Tokopedia, and Bukalapak). This experiment has been implemented in the web application using the Flask framework to generate products’ recommended items. This research employs the IMDb weight rating formula to get the best score lists and TF-IDF with Cosine similarity to create the similarity between products to produce related items.

Optimal Hybrid Filtering Strategy Using Adaptive Genetic-Fuzzy Logic Control for Harmonics Reduction in a Standalone Micro Hydroelectric Power Plant Coordinated with a PV System

Micro Hydro Power Plants are a type of power production that uses the force of river flows or waterfalls to generate electricity. The generator generates current waves and harmonic voltage, which are distorted wave disturbances that cause fundamental frequency multiplication. The major goal of this work is to design a reliable, efficient, and innovative harmonic mitigation approach for a stand-alone micro hydroelectric system that is coordinated with a photovoltaic renewable energy system utilising an active power filter. We may pick the active filter highest harmonic to be suppressed using the magnitude information supplied for each harmonic component. A hybrid filtering approach to remove harmonics and a novel MOGA optimization technique are part of the suggested harmonics reduction solution. The goal of this article is to determine the optimum filter for decreasing harmonics in an induction generator. As the harmonic damper, two filters were chosen: a passive filter and an active power filter. The suggested MOGA control method is compared to GA and evaluated on simulated data. In tracking harmonic components and fundamental frequency, the suggested MOGA control system provides high convergence speed and accuracy. It's extremely adaptable, and it can predict changes in the phase angle, amplitude, and fundamental frequency of harmonic components. When compared to the Genetic Algorithm method, it performs better. Simulation results using the SIMULINK/MATLAB simulation tool are delivered to evaluate the efficacy of the suggested active filter system. The impact of harmonic currents on the magnetic flux density is investigated using the rated condition as a reference. It has been established that the time harmonic is a significant element influencing generator performance. At the same time, the impacts of harmonic currents on the generator's eddy current loss, average torque, and torque ripple are investigated, as well as the mechanism of eddy current loss fluctuation.

A Hybrid Filtering Stage-Based Rotor Position Estimation Method of PMSM with Adaptive Parameter

The performance of sensorless control in a permanent magnet synchronous machine (PMSM) highly depends on the accuracy of rotor position estimation. Owing to its strong robustness, phase-locked loop (PLL) is widely used in rotor position estimation. However, due to the influence of harmonics existing in back electromotive force (EMF), estimation error occurs by using PLL. In this paper, a hybrid filtering stage-based PLL is proposed to improve the rotor position estimation. Adaptive notch filters and moving average filters are integrated together to eliminate harmonic EMF. To make the method effective under varying speed conditions, adaptive parameters design guidelines are provided, considering dynamic performance under a wide operating range. The proposed method can accurately detect rotor position even under harmonic EMF disturbances. It can also adjust the frequency adaptively based on the rotating speed of the rotor, which means the estimation performance is not deteriorated under rotating speed changing conditions. The simulation results verify the effectiveness of the proposed method.

A Novel Model for Explainable Hostel Recommender System Using Hybrid Filtering

Recommender systems help humans in filtering and finding the right information from the enormous amount of data. Hostels are more famous than hotels for solo travelers, but no prior research related to recommender systems has been conducted in this domain. Hostels allow users to provide multi-criteria ratings and traditional recommender systems are not able to provide effective recommendations in case of multi-dimensionality i.e. contextual information and multi-criteriaratings. So, we have proposed a novel hybrid recommender system (SAFCHERS) that chooses the hostel's features for computation dynamically and provides explainable and better recommendations than the traditional recommender systems.