Content-based Recommender System Research Articles

A cutting-edge artificial intelligence paradigm for entertainment-infused music recommendations

Personalized and fascinating music recommendations are becoming increasingly in demand as the advent of technology continues to change the way that people consume music. Traditional music recommendation systems primarily rely on user listening history and preferences, often neglecting the emotional and experiential dimensions that make music a deeply personal and entertaining endeavour. This research introduces a music recommendation system after classifying the tracks using a novel Gravitational Search Optimized Recursive Neural Networks (GS-RNN) approach. GS-RNN addressed this gap by integrating the Gravitational Search Algorithm (GSA) with Recursive Neural Networks (RNN) to create a content-based recommendation system that assesses auudio signal similarity. Evaluation metrics, including accuracy (80%), logarithmic loss (0.85), precision (84%), recall (83%), and F1-score (88%), demonstrate GS-RNN’s superiority over existing techniques. Genre-specific accuracy analysis further underscores the model’s capability to suggest songs within the same genre. Overall, GS-RNN presents a novel and effective paradigm for music recommendations.

Content Based Recommendation System on Netflix Data

After pandemic, OTT platforms are the most common platform to provide entertainment to users. Among all platforms, Netflix has become most the popular one. Data visualization of Netflix data can provide valuable insights and benefits in many ways like understanding viewer preferences, content optimization, personalized recommendation, quality and content performance evaluation, fraud detection to name a few. This research provides exploratory data visualization and provide a content based recommendation system on Netflix data as in real world applications, company uses these recommendation system algorithms to determine which system are better to improve users’ engagement of the platform.

Attributes Relevance in Content-Based Music Recommendation System

The possibility of recommendations of musical songs is becoming increasingly required because of the millions of users and songs included in online databases. Therefore, effective methods that automatically solve this issue need to be created. In this paper, the mentioned task is solved using three basic factors based on genre classification made by neural network, Mel-frequency cepstral coefficients (MFCCs), and the tempo of the song. The recommendation system is built using a probability function based on these three factors. The authors’ contribution to the development of an automatic content-based recommendation system are methods built with the use of the mentioned three factors. Using different combinations of them, four strategies were created. All four strategies were evaluated based on the feedback score of 37 users, who created a total of 300 surveys. The proposed recommendation methods show a definite improvement in comparison with a random method. The obtained results indicate that the MFCC parameters have the greatest impact on the quality of recommendations.

A content-based recommender system using stacked LSTM and an attention-based autoencoder

Recommender systems (RS) are popular in many areas, such as movies, music, news, books, research articles, search queries, and social tagging. Proprietary recommender systems are used by e-commerce websites like eBay, Amazon, and Alibaba to better match customers with products they are likely to purchase. This study suggests a recommender system that combines stacked long short-term memory (LSTM) and an attention-based autoencoder. This system would be used in a self-supervised learning paradigm, and the Amazon product datasets were used to run simulations. The results showed that the proposed method is more accurate, uses less computing power, and can be used on a large scale. No matter how big the data is, it can capture low-dimensional fixed latent representations and use the bare minimum of information already in the items, whether they are new or old. Several evaluation metrics show that the method works and that the cold start problem has been solved.

Towards Hyper-Relevance in Marketing: Development of a Hybrid Cold-Start Recommender System

Recommender systems position themselves as powerful tools in the support of relevance and personalization, presenting remarkable potential in the area of marketing. The cold-start customer problematic presents a challenge within this topic, leading to the need of distinguishing user features and preferences based on a restricted set of transactional information. This paper proposes a hybrid recommender system that aims to leverage transactional and portfolio information as indicating characteristics of customer behaviour. Four independent systems are combined through a parallelised weighted hybrid design. The first individual system utilises the price, target age, and brand of each product to develop a content-based recommender system, identifying item similarities. Secondly, a keyword-based content system uses product titles and descriptions to identify related groups of items. The third system utilises transactional data, defining similarity between products based on purchasing patterns, categorised as a collaborative model. The fourth system distinguishes itself from the previous approaches by leveraging association rules, using transactional information to establish antecedent and precedence relationships between items through a market basket analysis. Two datasets were analysed: product portfolio and transactional datasets. The product portfolio had 17,118 unique products and the included 4,408,825 instances from 2 June 2021 until 2 June 2022. Although the collaborative system demonstrated the best evaluation metrics when comparing all systems individually, the hybridisation of the four systems surpassed each of the individual systems in performance, with a 8.9% hit rate, 6.6% portfolio coverage, and with closer targeting of customer preferences and smaller bias.

ConCollA - A Smart Emotion-based Music Recommendation System for Drivers

Music recommender system is an area of information retrieval system that suggests customized music recommendations to users based on their previous preferences and experiences with music. While existing systems often overlook the emotional state of the driver, we propose a hybrid music recommendation system - ConCollA to provide a personalized experience based on user emotions. By incorporating facial expression recognition, ConCollA accurately identifies the driver’s emotions using convolution neural network(CNN) model and suggests music tailored to their emotional state. ConCollA combines collaborative filtering, a novel content-based recommendation system named Mood Adjusted Average Similarity (MAAS), and apriori algorithm to generate personalized music recommendations. The performance of ConCollA is assessed using various evaluation parameters. The results show that proposed emotion-aware model outperforms a collaborative-based recommender system.

Movie Popularity and Target Audience Prediction Using the Content-Based Recommender System

Movies are an important part of our daily entertainment, with the global movie industry experiencing significant growth and capturing the attention of people of all ages. However, only a few movies achieve success, leading to pressure on movie production stakeholders. Therefore, researchers and moviemakers require expert systems that can accurately predict the probability of a movie's success prior to its production. Most research on predicting movie popularity has focused on post-production stages, but it's essential to predict a movie's success at an early stage to enable necessary changes to be made. To this end, a content-based movie recommendation system (RS) has been proposed that uses preliminary movie features such as genre, cast, director, keywords, and movie description. This RS output and movie rating and voting information are then used to create a new feature set, which is input into a CNN deep learning (DL) model to build a multiclass movie popularity prediction system. The study also proposes a system to predict the popularity of the upcoming movie among different age groups, dividing the audience into four categories: junior, teenage, mid-age, and senior. The study uses publicly available data from the Internet Movie Database (IMDb) and The Movie Database (TMDb). The multiclass classification model implemented in this study achieved 96.8% accuracy, outperforming all benchmark models. Overall, this study highlights the potential of predictive and prescriptive data analytics in information systems to support industry decision- making.

Unveiling Anime Preferences: A Data-driven Analysis using MyAnimeList API

This research paper presents a data-driven analysis of anime preferences using the MyAnimeList (MAL) API. The objective of this study is to uncover valuable insights into the preferences of anime viewers by analyzing data obtained through MAL API. The paper begins with a comprehensive background that includes a detailed literature survey to understand anime recommendation systems. The methodology employed in this study involves utilizing the MAL API to collect and analyze data related to anime. The API provides access to a wide range of anime related information including ratings, ranks, genres etc. The collected data is then preprocessed to ensure its quality and suitability for analysis, involving data cleaning and transformation. Various data analysis techniques are applied to the preprocessed data. These techniques help identify patterns and trends in different fields such as genres, studios, media type etc. The findings from the data analysis provide valuable insights into the preferences of anime viewers and help to identify features that may be crucial while building the model. Then using cosine similarity model, content-based anime recommendation system is built after feature engineering. The results of this research highlight the importance of leveraging data-driven approaches in understanding and catering to the diverse preferences of anime viewers. By EDA uncovering patterns and trends in anime preferences, anime recommendation systems can be enhanced to provide more relevant recommendations to viewers. The implications of this research extend to the anime industry, where a deeper understanding of viewer preferences can inform content production and distribution strategies. In conclusion, this research paper presents a data-driven analysis and content-based recommendation system for anime using the MAL API and data science. The methodology employed involves data collection, preprocessing, Exploratory Data Analysis(EDA), Feature engineering and Model building. The findings contribute to enhancing anime recommendation systems and understanding the diverse preferences of anime viewers.

Semantics aware intelligent framework for content-based e-learning recommendation

E-learning accounts for the emergence of re-skilling, up-skilling, and augmenting the traditional education system by providing knowledge delivery. The meaningful learning approach is based on a constructivist process for conceptually modeling an individual’s current and past knowledge or experience towards personalization. This research proposes a novel framework in which semantic analysis of e-content is combined with deep machine learning techniques into an e-learning Intelligent Content-Based Recommendation System (ICRS), with the goal of assisting learners in selecting appropriate e-learning materials. The system tackles textual e-content to extract representative terms and their mutual semantic relationships by which a structure of the context-based graph is developed. Thus, the e-content is semantically represented according to the learner’s terms that are expanded using the ConceptNet semantic network to represent the textual knowledge meaningfully. Furthermore, a new approach is proposed wherein more contextual and semantic information among concepts with graphs are combined to infer the relative semantic relations between terms and e-learning resources in order to build the semantic matrix. This new approach is utilized to generate learners’ semantic datasets used for the classification of available resources to enrich the recommendation. Here, four machine learning (ML) models and an augmented deep learning (DL) recommender model named LSTMM, which are developed for the ICRS framework. The models have been evaluated and compared using a user sequential semantic dataset. Our results show that the LSTMM performs better than others in terms of Accuracy and F1 Score of 0.8453 and 0.7731 respectively.

Fashion Recommender System (FRS): Image Based Engine for Personalized Outfit

The methods used to predict how highly a person will evaluate a product or a group of people are known as recommendation systems. Books, movies, restaurants, and other products can be among the things commonly recommended. Objects where people differ in their preferences matter a lot. For predicting the preferences 2 methods are used, one a content based approach that considers the characteristic of an item, and the other is a collaborative method that evaluates choices by taking into account previous user behaviour. In this paper, a system for recommending fashion items is proposed, one that will base its recommendations on the provided clothing images' styles. The upper body and lower body clothing images, as well as those of a human model, are the main focus of this work. We have used the fashion image dataset from Kaggle website. This paper presents an idea to develop a content based recommendation system that uses convolutional neural network model, ResNet-50.

Recommender Systems: Collaborative Filtering and Content-based Recommender System

There are three algorithms of recommender systems proposed by this paper, which are item collaborative filtering(itemCF), user collaborative filtering(useCF) and content-based recommender system(CBRS). The principal goal of this paper is to try to ascertain which algorithm has the highest precision, after training based on the same dataset. In accordance with the data we chose and ceaseless testing, we observe itemCF contains the most accurate rate. However, we theoretically and empirically conceive each algorithm owns different advantages and drawbacks, should be used in the specific circumstance.

Hybrid recommendation system combining collaborative filtering and content-based recommendation with keyword extraction

With the development of recommendation systems, large amount of information collected from e-commerce could help customers to find the potential interesting products. Collaborative filtering and content-based recommendation systems are two common recommendation systems. While collaborative filtering has the problem of cold-start, content-based recommendation system could not explore the potential interests of users. Hybrid system combining these two techniques could achieve better results. This paper applies hybrid recommendation methods to the Amazon food reviews and evaluate the results in the aspects of precision, recall, diversity and novelty. It is found that the weighted hybrid recommendation system combing 0.95 weight of collaborative filtering and 0.05 content-based recommendation system achieves a good precision and diversity.

Kişiselleştirilmiş Haber Tavsiye Sistemi

Recommendation Systems are the methods that suggest the next choices of the user in a predictable way, based on the preferences made by the user before. This method is become even more popular nowadays and it can be applied to any topic or field that needs future estimation evaluating the data at hand. It is a kind of information extraction study. Furthermore, the fact that Amazon receives about 35% of its revenue from referral systems is an indication of how important this method is. However, news recommendation system which is a similar application area, is not also widely used as others. In this study, it is aimed to design a news recommendation system by taking into account the sites the user enters, the words that they searched for and bookmarks. The machine learning model has been trained with a data set that includes news categories and news content in order to present the news to the user as interested. By giving the data from the user environment to the trained model, the found interested categories of the user is processed instantly by the RSS. These news selected from RSS are shown to the user in order of priority regarding the daily news agenda. The real user test showed impressive accuracy as 89%. This solution presents a content-based recommendation system as nature of the problem.

Content Based Movie Recommendation System

Abstract: Recommender System is a tool which helps users find the required content and overcome information overload. It predicts interests of users by using Machine Learning algorithms and makes recommendation according to the interest of users. The primary content-based recommender system is the continuation and development of collaborative filtering, which does not need the user’s appraisal for items. Instead, the similarity is calculated based on the data of items that are selected by users, and then make the recommendation appropriately. With the augmentation of machine learning, the current content-based recommender system can build profile for users and products respectively. Building or renewing the profile according to the perusal of items that are bought or seen by users. The system can differentiate the user and the profile of items and then recommend the most resembling products. So, this recommender method that compel user and product directly can’t be brought into collaborative filtering model. The groundwork of content-based algorithm is acquisition and quantitative analysis of the content. The research of acquisition and filtering of text information are fully fledged, many current modified content-based recommender systems make recommendations according to the analysis of text data. This paper introduces content-based recommendation system for the movie websites. There are a lot of factors extracted from the movie, they are diverse and unique, which is also different from other recommender systems. We use these aspects to construct movie model and calculate similarity. We introduce a new outlook for setting weight of features, which improvises the representation of movie recommendations. Finally, we evaluate the approach to illustrate the improvement

Content-based Recommender System for Answering Web Based User Queries

Abstract: Online the idea of this paper is to give insights about a web-based project that aims to resolve the users query using machine learning technique of content-based filtering. The crux is to define the scope of the website. It becomes crucial to understand the business prospects of the project along with the relevance in the current technical world. As the paper goes about, various other papers and ideas will also be discussed that have been previously published and implemented. This paper will also break down the concept of content-based filtering for any individual who is a beginner in the field of machine learning.

Content-Based Movie Recommendation System Using MBO with DBN

The content-based filtering technique has been used effectively in a variety of Recommender Systems (RS). The user explicitly or implicitly provides data in the Content-Based Recommender System. The system collects this data and creates a profile for all the users, and the recommendation is generated by the user profile. The recommendation generated via content-based filtering is provided by observing just a single user’s profile. The primary objective of this RS is to recommend a list of movies based on the user’s preferences. A content-based movie recommendation model is proposed in this research, which recommends movies based on the user’s profile from the Facebook platform. The recommendation system is built with a hybrid model that combines the Monarch Butterfly Optimization (MBO) with the Deep Belief Network (DBN). For feature selection, the MBO is utilized, while DBN is used for classification. The datasets used in the experiment are collected from Facebook and MovieLens. The dataset features are evaluated for performance evaluation to validate if data with various attributes can solve the matching recommendations. Each file is compared with features that prove the features will support movie recommendations. The proposed model’s mean absolute error (MAE) and root-mean-square error (RMSE) values are 0.716 and 0.915, and its precision and recall are 97.35 and 96.60 percent, respectively. Extensive tests have demonstrated the advantages of the proposed method in terms of MAE, RMSE, Precision, and Recall compared to state-of-the-art algorithms such as Fuzzy C-means with Bat algorithm (FCM-BAT), Collaborative filtering with k-NN and the normalized discounted cumulative gain method (CF-kNN+NDCG), User profile correlation‑based similarity (UPCSim), and Deep Autoencoder.

Generative Adversarial Network-based Visual Similarity Recommendation

The purpose of a content-based recommendation system is to rank a set of results by how closely they match the user’s query. These days, virtually every online marketplace has some kind of recommendation system in place to steer customers to more goods and services they might enjoy. In this piece, we’ll go through how we used Generative Adversarial Networks to build an image retrieval system for e-commerce sites, with a focus on locating the best possible shoe photos to go along with a given product image. The writers of this paper conducted the study presented here. We demonstrate the outcomes of our study of the proposed deep learning network’s performance on a benchmark dataset, and we compare the network’s results to those of state-of-the-art systems.

Identifying Fake Digital Information Using Machine Learning Algorithms: Performance Analysis and Recommendation System

This work focuses on the detection of fake digital information using various machine learning and deep learning algorithms to prevent its spread through Internet of Things (IoT) devices and systems. The research highlights the significance of detecting and preventing false or misleading information in critical areas such as healthcare, public safety, and emergency response. The study compares the performance of several supervised machine learning algorithms and identifies logistic regression as the most accurate (98.03%). The empirical analysis used data from The Indian Express, PolitiFact, and Kaggle and leveraged natural language processing (NLP) to prepare, clean, and model the data. To detect fraudulent posts, the study employed random forest, a supervised machine learning algorithm, which achieved an impressive accuracy rate of 99.71% on a Kaggle dataset. The research also developed a model for detecting false reporting related to COVID-19, utilizing the support vector machine technique, which achieved an accuracy rate of 78.69%. The presented work also determined the authenticity of images through convolutional neural networks (CNNs). Lastly, a content-based recommendation system was developed to enhance people’s security and confidence.

Content-based Recommender System with Descriptive Analytics

Content-based Recommender System with Descriptive Analytics

A content-based recommender system with consideration of repeat purchase behavior

With the increasing popularity of online shopping and information explosion, personalized recommender systems for e-commerce become more and more necessary, which helps customers find the desired products efficiently among variety of categories based on their previous behavior such as buying pattern and rating history. However, most recommender systems for e-commerce adopt binary (purchase/non-purchase) or subjective weighting methods to represent the customer preferences, which is hard to predict their profiles precisely since rapid change in tastes. Therefore, this study focuses on the application of transactional data. A personalized recommender system for e-commerce (PROSE) is proposed in order to enhance the quality of recommendations by integrating the architecture of traditional content-based recommender system with a new component called feedback adjuster, which is designed to make customer implicit feedback reflects the reality of preferences as possible through taking into consideration their behavior of repeat purchase. The computational results indicate that the proposed algorithm is able to outperform other algorithms.