Abstract
This Gym Activities Recommender System serves to upgrade gym sessions by developing custom workout suggestions suited for each user's tastes as well as fitness objectives. The majority of gym members including newcomers battle to find appropriate exercises because they lack directional support and experience exercise complexity. The absence of proper guidance leads users to experience diminished motivation and choose wrong exercises that results in futile workouts. The proposed recommendation system bases its operation on Content-Based Filtering (CBF) to process metadata from different gym exercises which produces personalized workout recommendations. User-provided fitness objectives along with choice of workout exercises and experience background help the system develop customized workout profiles. The system matches users with appropriate exercises based on two similarity calculation methods which include cosine similarity alongside TF-IDF (Term Frequency-Inverse Document Frequency). The research adopts a formal methodology which combines gym activity dataset compilation and systematic design of the system with algorithm development and performance assessment. The recommendation system achieves performance evaluation through measurements of accuracy with 81.33% and precision with 81.66% as well asrecall 100% and F1-score with 89%.The implementation of machine learning algorithms in content-based filtering methods delivers better gym activity recommendations which enhances usersatisfaction as well as engagement. Users experience simplified workout selection through the system because it provides matched recommendations that boost their fitness development. The research demonstrates why artificial intelligence needs to enter fitness applications for data-oriented user-focused workout planning that enhances both workout adherence and health results. TABLE OF CONTENT CONTENT PAGE SUPERVISOR APPROVAL i STUDENT DECLARATION i ACKNOWLEDGEMENT ii ABSTRACT iii TABLE OF CONTENT iv TABLE OF FIGURES viii LIST OFTABLES x LIST OFABBREVIATIONS xi CHAPTER 1 1 1.1 Background study 1 1.2 Problem statement 2 1.3 Objective 3 1.4 Project scope 4 1.5 Project significant 5 1.6 Research framework 6 1.7 Conclusion 7 CHAPTER 2 8 2.1 Introduction 8 2.2 Recommendation system 8 2.2.1 The use of recommendations system 9 2.2.3 Recommendation system technique 11 2.2.4 Advantages and disadvantages of recommendation design techniques 15 2.3 Gym activitiesrecommendation 20 2.3.1 Problem with gym activities recommendation 20 2.3.2 Benefit of gym activities existence 21 iv 2.3.3 The recommendation in gym activities 21 2.4 Similarity computing 22 2.4.1 Cosine Similarity 22 2.4.2 Term Frequency-Inverse Document Frequency (TF-IDF) 23 2.5 Implementation of Content based filtering in various problem 24 2.6 The implications of gym activity recommender 32 2.7 Implication of Literature Review 43 2.8 Conclusion 44 CHAPTER 3 45 3.1 Overview of Research Methodology 45 3.1.1 Detailed of Research Framework 46 3.2 Preliminary Phase 48 3.2.1 Literature Study 48 3.2.2 Data pre-processing 48 3.2.3 Data description 51 3.3 Design Phase 52 3.3.1 System Architecture 53 3.3.2 Flowchart 54 3.3.3 User Interface Design 55 3.3.4 Pseudocode of SelectedAlgorithm 57 3.4 Model Training 58 3.5 Model Testing 59 3.6 Performance Evaluation 60 3.6.1 Recall & Precision 60 3.6.2 F-Measure 60 3.6.3 Accuracy 61 3.7 Prototype Implementation 61 v
Showing the abstract — retrieve the full paper via the Exa API.