Academia.eduAcademia.edu

Figure 2 – uploaded by RAMNATH M . GAIKWAD

See full PDFdownloadDownload figure
Fig. 2. Support Vector Machines are explained in detail [23] Support Vector Machines (SVMs) are classification algorithm that represents data in an n-dimensional space and uses classifier lines to separate and graph the data. SVM is a contemporary predictor approach applicable to both classification and regression [23].

Figure 2 Support Vector Machines are explained in detail [23] Support Vector Machines (SVMs) are classification algorithm that represents data in an n-dimensional space and uses classifier lines to separate and graph the data. SVM is a contemporary predictor approach applicable to both classification and regression [23].

Related Figures (6)

Fig. 1. Different levels of sentiment analysis [25]. B. Sentiment Analysis at the Sentence Level
Fig. 1. Different levels of sentiment analysis [25]. B. Sentiment Analysis at the Sentence Level
Fig. 4. Architecture of the proposed model. This study proposes a methodological framework for conducting sentiment analysis on Marathi text data. The methodology consists of several key steps aimed at effectively preparing, modeling, and evaluating the data for sentiment analysis. “Fig. 4”, illustrates the architecture of the proposed model designed for Marathi text sentiment analysis.
Fig. 4. Architecture of the proposed model. This study proposes a methodological framework for conducting sentiment analysis on Marathi text data. The methodology consists of several key steps aimed at effectively preparing, modeling, and evaluating the data for sentiment analysis. “Fig. 4”, illustrates the architecture of the proposed model designed for Marathi text sentiment analysis.
Fig. 3. Basic Recurrent Neural Network (RNN) architecture
Fig. 3. Basic Recurrent Neural Network (RNN) architecture
\BLE III. ABOVE THE TABLE ARE SENTIMENT LABELS AND SCORES. After preprocessing, the final dataset for model training was refined to eliminate duplicates, resulting in the removal of 130 instances from the initial count of 3011. The processed data now comprises 2882 unique entries. Each text entry is associated with a sentiment label, categorizing it as being positive, negative, or neutral, along with a corresponding sentiment score. Among the unique entries, the most frequent sentiment label is " sua " (Effronted), classified as negative sentiment, with a sentiment score of 1.
\BLE III. ABOVE THE TABLE ARE SENTIMENT LABELS AND SCORES. After preprocessing, the final dataset for model training was refined to eliminate duplicates, resulting in the removal of 130 instances from the initial count of 3011. The processed data now comprises 2882 unique entries. Each text entry is associated with a sentiment label, categorizing it as being positive, negative, or neutral, along with a corresponding sentiment score. Among the unique entries, the most frequent sentiment label is " sua " (Effronted), classified as negative sentiment, with a sentiment score of 1.
We employed google translate to develop a marathi dictionary, as shown in “Table I’, we used the English resource senticnet dictionary. Using Google Translate, we will develop a Marathi dictionary.
We employed google translate to develop a marathi dictionary, as shown in “Table I’, we used the English resource senticnet dictionary. Using Google Translate, we will develop a Marathi dictionary.
ABLE II. THE END OUTCOME OF THE ENGLISH TO MARATHI TRANSLATION PROCEDURE. The list was translated from English to Marathi using Google Translate, as show in “Table II”, marathi experts then reviewed it, removing multi-word entries, mistranslations, and unclear terms. However, translating ‘personality,’ 'English SenticNet,' and 'SentiWordNet'’ posed challenges. Google Translate was utilized due to its speed, cost-effectiveness, and accuracy.
ABLE II. THE END OUTCOME OF THE ENGLISH TO MARATHI TRANSLATION PROCEDURE. The list was translated from English to Marathi using Google Translate, as show in “Table II”, marathi experts then reviewed it, removing multi-word entries, mistranslations, and unclear terms. However, translating ‘personality,’ 'English SenticNet,' and 'SentiWordNet'’ posed challenges. Google Translate was utilized due to its speed, cost-effectiveness, and accuracy.
Related topics:
Machine LearningSentiment AnalysisAutomatic Speech Recognition
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved