Análisis De Datos Masivos De Usuarios En Redes Sociales Mediante Técnicas De Segmentación De Usuarios Y Predicción De Comportamientos
This thesis, titled “Analysis of Massive User Data on Social Media through Segmentation and Behavior Prediction Techniques”, addresses the need to understand and anticipate interaction patterns on digital platforms by analyzing large volumes of data. The research focuses on the development of a mode...
محفوظ في:
| المؤلف الرئيسي: | |
|---|---|
| التنسيق: | masterThesis |
| اللغة: | spa |
| منشور في: |
2025
|
| الموضوعات: | |
| الوصول للمادة أونلاين: | http://dspace.unach.edu.ec/handle/51000/15667 |
| الوسوم: |
إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
|
إضافة وسم | الملخص: | This thesis, titled “Analysis of Massive User Data on Social Media through Segmentation and Behavior Prediction Techniques”, addresses the need to understand and anticipate interaction patterns on digital platforms by analyzing large volumes of data. The research focuses on the development of a model based on segmentation techniques and machine learning to predict user engagement with audiovisual content. To achieve this objective, a dataset from the Kaggle repository, specifically the “YouTube Dislikes Dataset” by Dmitry Nikolaev, was utilized. This dataset contains quantitative metrics associated with YouTube videos, such as views, likes, dislikes, comment count, and publication date. The adopted methodology followed the CRISP-DM (Cross Industry Standard Process for Data Mining) framework, which structured the process into stages including data understanding, preparation, modeling (using algorithms like Random Forest), and result evaluation. Additionally, feature engineering and class balancing techniques were applied to optimize the predictive model's performance. The results show that variables such as the engagement ratio and views per day are key determinants in classifying videos with high engagement. The model achieved an accuracy exceeding 96%, demonstrating the effectiveness of the applied approach. In conclusion, this study shows that the use of machine learning techniques on massive social media data allows not only the identification of digital behavior patterns but also the development of predictive tools with practical applications for decision-making in content platforms and digital marketing. |
|---|