International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 10 Issue: 06 | June 2023
p-ISSN: 2395-0072
www.irjet.net
The Identification of Depressive Moods from Twitter Data by Using Convolutional Neural Network with Text Data along with Emoji Pratibha M. Jadhav1, Dr. Sonia2, Dr. Anjali N. Kulkarni3 1 Research Scholar, Computer Science, JJT University Chudela, Rajasthan and Assistant Professor, Department of
Computer Science, C.K. Thakur Arts, Commerce and Science College, New Panvel, India
2Associate Professor, Department of Computer Science, JJT University Chudela, Rajasthan, India 3 Assistant Professor, C.K. Thakur Arts, Commerce and Science College, New Panvel, India
---------------------------------------------------------------------***--------------------------------------------------------------------Machines (SVMs) and Multinomial Naive Bayes (MNBs) for Abstract - The identification of depressive moods from social
microblog sentiment classification [3], while Bifet and Frank (2010) assessed sentiment analysis algorithms using WEKA and MOA software [4]. Other methods include automated corpus collection [1], Kouloumpis (2011) [13] utilized machine learning algorithms and lexicon-based methods for sentiment analysis on Twitter data. The study highlighted the challenges of performing sentiment analysis on Twitter due to the platform’s unique characteristics and demonstrated the feasibility of such an analysis. Similarly, Barbosa (2010) [2] proposed a robust sentiment detection approach that incorporated domain-independent features and employed a machine learning algorithm, showing promise in overcoming the biased and noisy nature of Twitter data. Liu K.L. (2018) [16] sought to improve Twitter sentiment analysis by incorporating emoticons into language models, successfully enhancing the accuracy of sentiment classification.
media platforms like Twitter has gained significant attention in recent years. But very little research is there on emoji sentiment analysis. In this research paper, we propose a Convolutional Neural Network (CNN) model that leverages both text data and emoji representations for accurate identification of depressive moods in Twitter data. The model is developed using popular Python libraries, including pandas, scikit-learn, TensorFlow's Keras, and NLTK. The performance of the CNN model is evaluated using metrics such as accuracy, precision, recall, and F1-score. Additionally, the paper explores the integration of emoji representations to enhance the detection of depressive moods. Key Words: CNN, depressive, scikit-learn, NLTK, emoji
1. INTRODUCTION Depression is a significant mental health concern affecting a large portion of the population. Early detection and intervention are crucial for improving patient outcomes. Twitter is a social media platform that allows users to share short, text-based messages with other users. Twitter has become one of the most popular social media platforms, with over 320 million active users. Unlike other social media platforms, Twitter is primarily text-based. Twitter is also known for its heavy use of hashtags, which are used to categorize tweets and make them easier to search for. Twitter is often used to share news and information, as well as personal thoughts and opinions. Textual content, including emojis, can provide valuable insights into an individual's emotional state. This research paper aims to compare the effectiveness of CNN models trained on text data combined with emoji representations in predicting depression. By leveraging the expressive power of CNNs and incorporating emoji sentiments, the models aim to improve accuracy in depression prediction.
2.2 Sentiment Analysis in HealthCare Applying sentiment analysis to healthcare can provide valuable patient insights, support disease prediction, and monitor treatment efficacy. Paul M. (2011) analyzed social media data to identify health trends and predict disease outbreaks [17]. Research specifically dedicated to sentiment analysis for mental health applications has flourished in recent years. Similarly, De Choudhury et al. (2013) used social media to predict the onset of depression in individuals [18]. 2.3 Emoji Analysis and Emotion Detection Emoticons play a significant role in conveying emotions in text-based communication. C.Yh. Chang (2017) studied asynchronous web-based peer responses in an English writing class using text-based emoticons, finding that emoticons were predominantly used in positive contexts [7]. The research by Francesco Barbieri (2017) predicted which emojis would be used based on the words in text-based tweets [10]. Studies by Bhavesh Tupkar (2021) [5] and Gupta S. (2023) [11] have further validated the influence of emojis on sentiment polarity in tweets. Joao Miguel (2018) introduced Emojinating, a technique that aids in brainstorming sessions by generating new emojis. This
2. RELATED WORK 2.1 Sentiment Classification Techniques Numerous classification techniques exist for sentiment analysis. Bermingham et al. (2010) utilized Support Vector
© 2023, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 1257