ارسل ملاحظاتك

ارسل ملاحظاتك لنا







Review in Analyzing Social Media Sentiment through Visual Insights Using Machine Learning

المصدر: مجلة الدراسات المستدامة
الناشر: الجمعية العلمية للدراسات التربوية المستدامة
المؤلف الرئيسي: Ibrahim, Rasha Mahmoud (Author)
مؤلفين آخرين: Abdulbaqi, Huda Abdulaali (Co-Author)
المجلد/العدد: مج6, ملحق
محكمة: نعم
الدولة: العراق
التاريخ الميلادي: 2024
التاريخ الهجري: 1446
الشهر: آب
الصفحات: 786 - 797
ISSN: 2663-2284
رقم MD: 1482753
نوع المحتوى: بحوث ومقالات
اللغة: الإنجليزية
قواعد المعلومات: EduSearch
مواضيع:
كلمات المؤلف المفتاحية:
Sentiment Analyses | Machine Learning | Image | Text | Features Vector | X- Dataset
رابط المحتوى:
صورة الغلاف QR قانون

عدد مرات التحميل

1

حفظ في:
المستخلص: The online platform for social interaction provides insight into the psychological behavior of people and helps analyze the collective viewpoint on social and political issues. In order to determine the opinions, feelings, and subjectivity of text, images, and other expressions, this field of study employs a computational approach. The realm of visual-textual sentiment analysis, the objective is to forecast sentiment by leveraging both image and text data. The primary hurdle in this area lies in procuring robust visual features that contribute to accurate sentiment prediction, given the varied and diverse characteristics often found in input images. The classification algorithm employed determines the variation in emotion polarity. Depending on the techniques and algorithm employed, the same text can be classified as neutral by certain classifiers, positive by others, and negative by additional methods. The datasets used can be divide into psychological-based or social media-based categories. Feature selection is important for data analysis, as data features encode the information for the system to infer correlations between visual features and sentiment. There are three general approaches to Sentiment Analysis: text-based, image based, and multi-modal representation learning combining visual and textual information through multi-modal embedding systems.

ISSN: 2663-2284

عناصر مشابهة