Human Annotation and Interpretation of Public Sentiments about Jio Coin marked in Social Networks using Machine Learning Algorithms

Emotions have a large influence on product loyalty and found that an individual have a positive emotional relationship with identified product, trust in the company and make purchases. The customer's decision is not founded on their facts/ information, but they rely on their emotions to get brand loyalty decisions. However, companies are struggling to adapt this strategy even with the huge amount of data they acquired, their utilization of this data as marketing influence and ongoing personalization tactics is limited due to the problem of analyzing voluminous data. This work offers new approach for automated categorization of emotions from textual data and was based on a sentence level utilizing the language translation machine and applying of learning machines and ensemble models. The experiment reveals that the proposed automatic emotion classifier performs well in classifying angry comments both in training and testing evaluation periods. Furthermore, the proposed emotion classifier produces a result of 75% accuracy under ensemble Naïve Bayes (NB) in the classification of sentence emotion during training period. Although during the actual testing period of the test data set, the proposed system was able to correctly classify 76% emotions of the sentences. Some of the reasons and causes for misclassification are analyzed and presented in this paper.

Section: Self-labeled Emotion Manual Labeled Data Creationmentioning

confidence: 99%

Emotion Recognition from Blog Comments Based Automatically Generated Datasets and Ensemble Models

Patacsil¹

2020

“…We have adapted the lexicon based approach for classifying sentiments with the concept of Word Sense Disambiguation using WordNet and SentiWordNet lexical sources [14,15]. The words of English Language are grouped into set of synonyms called synset, in the lexical resource WordNet.…”

Section: Classifying Sentimentsmentioning

confidence: 99%

Public Sentiment Insights Analysis using Word Sense Disambiguation Application on Twitter Data during a Pandemic – Covid’19

Anuratha¹,

Parvathy²,

Priyadarshini³

et al. 2020

The entire world is affected because of the global pandemic Covid-19 due to the virus belongs to the family of Coronavirus. As the spread of infection and mortality rate is rapid people have started developing assorted emotions about the crisis. It is more significant to administer the mental health and Psychological wellbeing of public during a crisis like this. As many of the people broadly use the social media like twitter for sharing their opinions and thoughts, our work utilizes the Covid specific Tweets posted by the Tweeple and analyse them to understand the sentiments exhibited regarding the situation. After the tweets are collected and the real sentiments behind them are discovered using the classifier model developed using the Machine Learning methods. The experimental results may be used by the respective authorities to take necessary initiatives for addressing the concerns that affect the wellbeing of the society and the economic wellbeing as well. As our word uses Lexical based sentiment analysis, it is important to remove the ambiguities of a word which is a main challenge of this technique on sentiment analysis. To improve the performance of the Sentiment Analysis we have used the lexical dictionaries Wordnet and SentiWordNet along with Word Sense Disambiguation (WSD) to detect and remove the ambiguities understanding the context of the term used in the tweets.

“…All systems undergo standard pre-processing stages described in [27]. The only difference, Luganda dataset materials used for Morpho SMT systems, first undergo morphological segmentation based on compiled GNC prefixes, after tokenization during standard pre-processing [28], [29]. Word alignment and language model training for the systems follows using Giza++ and KenLM respectively, setting our language model to 5 n-gram, sentence length limit at 90 throughout all the systems.…”

Section: Luganda Smt Models Trainedmentioning

confidence: 99%

English to Luganda SMT: Ganda Noun Class Prefix Segmentation for Enriched Machine Translation

Ssentumbwe¹

2019

Luganda or Ganda is a morphologically rich and low-resource language from Uganda. The morphological richness of Luganda sentences has an impact on the quality of translation and this work looks at improving machine translation (MT) for English to Luganda. Luganda sentence formation bases on 10 noun classes with a prefix for singular and plural. In various aspects, the interaction of these class prefixes in sentences usually enforces different words such as nouns, verbs, adverbs and adjectives to be in agreement with the subject in a given sentence. Such scenarios have resulted in various Luganda word inflectional and derivational tendencies because each noun, verb, adverb or adjective finds itself having to combine with a prefix from a class it belongs and thus creating new word forms. Using 6 statistical machine translation (SMT) models divided equally into base and morphology models, we propose a procedure to segment Luganda sentences from our English-Luganda parallel Bible corpus. Our morphological segmentation approach bases on Ganda Noun Class (GNC) prefixes and we design a tool we call GandaKIT to segment Luganda sentences at pre-processing stage and desegement them after translation at post processing stage. In experiments, we compare translation performance of SMT base models against systems trained with morphological segmentation at pre-processing stage. Our results show an improvement in MT performance over base models ranging by a difference of 1.58 BLEU points and 0.2257 NIST score for our best system.