Sentiment analysis of the Hamas-Israel war on YouTube comments using deep learning Scientific Reports

It can gradually label instances in the order of increasing hardness without the requirement for manual labeling effort. Since then, GML has been also applied to the task of aspect-level sentiment analysis6,7. It is worthy to point out that as a general paradigm, GML is potentially applicable to various classification tasks, including sentence-level sentiment analysis as shown in this paper. Even though the existing unsupervised GML solutions can achieve competitive performance compared with many supervised approaches, without exploiting labeled training data, their performance is still limited by inaccurate and insufficient knowledge conveyance.

Therefore naturally, the most successful approaches are using supervised models that need a fair amount of labelled data to be trained. Providing such data is an expensive and time-consuming process that is not possible or readily accessible in many cases. Additionally, the output of such models is a number implying how similar the text is to the positive examples we provided during the training and does not consider nuances such as sentiment complexity of the text. This section analyses the performance of proposed models in both sentiment analysis and offensive language identification system by examining actual class labels with predicted one.

The Bidirectional-LSTM layer receives the vector representation of the data as an input to learn features once the data has been preprocessed and the embedding component has been constructed. Bi-directional LSTM (Bi-LSTM) can extract important contextual data from both past and future time sequences. Bi-LSTM, in contrast to LSTM, contains forward and backward layers for conducting additional feature extractions which is suitable for Amharic language because the language by its nature needs context information to understand the sentence.

After working out the basics, we can now move on to the gist of this post, namely the unsupervised approach to sentiment analysis, which I call Semantic Similarity Analysis (SSA) from now on. In this approach, I first train a word embedding model using all the reviews. The characteristic of this embedding space is that the similarity between words in this space (Cosine similarity here) is a measure of their semantic relevance.

The analysis uses advanced algorithms and natural language processing (NLP) to evaluate the emotions behind social media interactions. RNNs, including simple RNNs, LSTMs, and GRUs, are crucial for predictive tasks such as natural language understanding, speech synthesis, and recognition due to their ability to handle sequential data. Therefore, the proposed LSTM model classifies the sentiments with an accuracy of 85.04%. To experiment, the researcher collected a Twitter dataset from the Kaggle repository26. Therefore, their versatility makes them suitable for various data types, such as time series, voice, text, financial, audio, video, and weather analysis.

Performance evaluation

Instead of creating dozens of short, disparate pages, each with its own topic, consider creating “ultimate guides” and more comprehensive resources that your users will find valuable. Create content that clearly and concisely answers a common query at the top of the page before delving into more specific details. This all helps Google in its goal to provide a better experience for its users by delivering quality and giving preference to relevant content results. Instead of answering “How big is a blue whale,” Google would seek to match the specific keywords from the phrase “How big is it? Many things have changed since 2010 when SEO was more concerned with getting as many backlinks as you could and including as many keywords as possible.

Moreover, it also plays a crucial role in offering SEO benefits to the company. Upon parsing, the analysis then proceeds to the interpretation step, which is critical for artificial intelligence algorithms. For example, the word ‘Blackberry’ could refer to a fruit, a company, or its products, along with several other meanings.

Research methodology

The first sentence is an example of a Positive class label in which the model gets predicted correctly. The same is followed for all the classes such as positive, negative, mixed feelings and unknown state. Affective computing and sentiment analysis21 can be exploited for affective tutoring and affective entertainment or for troll filtering and spam detection in online social communication. In recent years, classification of sentiment analysis in text is proposed by many researchers using different models, such as identifying sentiments in code-mixed data9 using an auto-regressive XLNet model. Despite the fact that the Tamil-English mixed dataset has more samples, the model is better on the Malayalam-English dataset; this is due to greater noise in the Tamil-English dataset, which results in poor performance. These results can be improved further by training the model for additional epochs with text preprocessing steps that includes oversampling and undersampling of the minority and majority classes, respectively10.

Another potential approach involves using explicitly trained machine learning models to identify and classify these features and assign them as positive, negative, or neutral sentiments. These models can subsequently be employed to classify the sentiment conveyed within the text by incorporating slang, colloquial language, irony, or sarcasm. This facilitates a more accurate determination of the overall sentiment expressed. The work in20 proposes a solution for finding large annotated corpora for sentiment analysis in non-English languages by utilizing a pre-trained multilingual transformer model and data-augmentation techniques. The authors showed that using machine-translated data can help distinguish relevant features for sentiment classification better using SVM models with Bag-of-N-Grams.

In this article, we will develop a multi-class text classification on Yelp reviews using BERT. Interestingly, the best threshold for both models (0.038 and 0.037) was close in the test set. And at this threshold, ChatGPT achieved an 11pp better accuracy than the Domain-Specific model (0.66 vs. 077). Also, ChatGPT showed a much better consistency across threshold changes than the Domain-Specific Model. In summary, if you have thousands of sentences to process, start with a batch of a few half-dozen sentences and no more than 10 prompts to check on the reliability of the responses. Then, slowly increase the number to verify capacity and quality until you find the optimal prompt and rate that fits your task.

Besides, these language models are able to perform summarization, entity extraction, paraphrasing, and classification. NLP Cloud’s models thus overcome the complexities of deploying AI models into production while mitigating in-house DevOps and machine learning teams. These things are vital for SEO in an age of semantic search, where machine learning and natural language processing are helping search engines understand context and consumers better. Considering the practical difficulties connected with the clinical application of such tasks66, a possible way forward could be the combination of traditional speech elicitation tasks with corpus-based approaches67. The use of different speech elicitation tasks, capable of triggering a broader variety of linguistic uses, could also be important. We first analyzed media bias from the aspect of event selection to study which topics a media outlet tends to focus on or ignore.

Normalization brings each character in the designated uni-code array ( FF) for the Urdu dialect. This section contains the experimental description of applied machine learning, rule-based, deep learning algorithms and our proposed two-layer stacked Bi-LSTM model. These algorithms have been trained and tested on our proposed UCSA-21 corpus and UCSA50 datasets which are publically available. A research study focusing on Urdu sentiment analysis41 created two datasets of user reviews to examine the efficiency of the proposed model. Only 650 movie reviews are included in the C1 dataset, with each review averaging 264 words in length. You can foun additiona information about ai customer service and artificial intelligence and NLP. The other dataset named C2, contains 700 reviews about refrigerators, air conditions, and televisions.

It also developed an evaluating chatbot performance feature, which offers a data-driven approach to a chatbot’s effectiveness so you can discover which workflows or questions bring in more conversions. Additionally, Idiomatic has added a sentiment score tool that calculates the score per ticket and shows the average score per issue, desk channel, and customer segment. MonkeyLearn has recently launched an upgraded version that lets you build text analysis models powered by machine learning. It has redesigned its graphic user interface (GUI) and API with a simpler platform to serve both technical and non-technical users. Additionally, it has included custom extractors and classifiers, so you can train an ML model to extract custom data within text and classify texts into tags.

The performance of the trained models was reduced with 70/30, 90/10, and another train-test split ratio. During the model process, the training dataset was divided into a training set and a validation set using a 0.10 (10%) validation split. Therefore train-validation split allows for monitoring of overfitting and underfitting during training. The training dataset is used as input for the LSTM, Bi-LSTM, GRU, and CNN-BiLSTM learning algorithms. Therefore, after the models are trained, their performance is validated using the testing dataset.

Potential strategies include the utilization of domain-specific lexicons, training data curated for the specific cultural context, or applying machine learning models tailored to accommodate cultural differences. Integrating cultural awareness into sentiment analysis methodologies enables a more refined understanding of the sentiments expressed in the translated text, enabling comprehensive and accurate analysis across diverse linguistic and cultural domains. Alternatively, machine learning techniques can be used to train translation systems tailored to specific languages or domains.

• Initially, the weights of the similarity factors (whether KNN-based or semantic factors) are set to be positive (e.g., 1 in our experiments) while the weights of the opposite semantic factors are set to be negative (e.g., − 1 in our experiments).
• The batch size was increased from 64 to 100, and the epoch number was decreased from 10 to 9.
• If the model is trained based on not only words but also context, this misclassification can be avoided, and accuracy can be further improved.
• These pre-trained models are trained on large corpus in order to capture long-term semantic dependencies.
• You can click on each category to see a breakdown of each issue that Idiomatic has detected for each customer, including billing, charge disputes, loan payments, and transferring credit.

However, in spite of the progress, these methods often rely on manual observation and interpretation, thus inefficient and susceptible to human bias and errors. I chose frequency Bag-of-Words for this part as a simple yet powerful baseline approach for text vectorization. Frequency Bag-of-Words assigns a vector to each document with the size of the vocabulary in our corpus, each dimension representing a word.

This finding is consistent with the increases in negative sentiment observed across all parts of speech in both The Economist and Expansión. The data we used to carry out the test correspond to the frequency values of negative polarity in the total of adjectives, adverbs, nouns and verbs in Spanish and English extracted from the pre-covid and covid corpus (Table 6). One of the evident issues arising from the analysis of this corpus is that the frequencies of emotions are similar in number to those in the Spanish corpus.

As an audience member, I have grown accustomed to the current stasis of his art. I eagerly anticipate the day Wes Anderson allows himself to step outside the defining and restrictive genre of himself. When looking at Wes Anderson’s work we notice that there is a heavy reliance on the consistency of semantic criteria without the presence of syntactic narrative justification. This leads to weak overall narratives that lack the structure necessary to support and justify the ornate details of Anderson’s work.

Training the system on extensive datasets and employing specialized machine learning algorithms and natural language processing methodologies can enhance the accuracy of translations, thereby reducing errors in subsequent sentiment analysis. Although it demands access to substantial datasets and domain-specific expertise, this approach offers a scalable and precise solution for foreign language sentiment analysis. The results ChatGPT App presented in this study provide strong evidence that foreign language sentiments can be analyzed by translating them into English, which serves as the base language. This concept is further supported by the fact that using machine translation and sentiment analysis models trained in English, we achieved high accuracy in predicting the sentiment of non-English languages such as Arabic, Chinese, French, and Italian.

A few research employing deep learning, semantic graphs and multimodal based system (MBS) have been undertaken on the areas of emotion classification51, concept extraction52, and user behavior analysis53. A unique CNN Text word2vec model was proposed in the research study51 to analyze emotion in microblog texts. According to the testing results the suggested MBS52 has a remarkable ability to learn the normal pattern of users’ everyday activities and detect anomalous behaviors. SemEval challenges are the most prominent efforts taken in the existing literature to create standard datasets for SA.

If you do not do that properly, you will suffer in the post-processing results phase. For this subtask, the winning research team (i.e., which ranked best on the test set) named their ML architecture Fortia-FBK. Inspired by this competition’s discoveries, some colleagues and I made a research article (Assessing Regression-Based Sentiment Analysis Techniques in Financial Texts) where we implemented our version of Fortia-FBK and evaluated ways to improve this architecture. No significant associations were observed between the linguistic PCs and the cognitive and sociocognitive measures for participants in Cluster 1 (|rs|≤ .21, ps ≥ 0.185) (Fig. 5A). Conversely, Cluster 2 exhibited an overall stronger pattern of correlations between linguistic and cognitive aspects. No other robust significant associations between the linguistic-based PCs and the ToM PST subscores were found in Cluster 2.

To build a word representation of the data for the deep learning model the researcher employs Word2Vec as an embedding model. After preprocessing and converting the datasets to a format that can be analyzed, the words in the sentence must be represented as vectors so that Word2Vec can calculate similarity, analogy. The embedding layer converts the input what is semantic analysis into an \(N\times M\) dimensional vector, where N represents the longest sentence in the dataset and M represents the embedding dimension. In this study, the selection of deep learning models was contingent on their suitability for Amharic sentiment analysis. During the model selection process criteria that is noted by Refs.22,23,24 were considered.

ChatGPT, in its GPT-3 version, cannot attribute sentiment to text sentences using numeric values (no matter how much I tried). However, specialists attributed numeric scores to sentence sentiments in this particular Gold-Standard dataset. The last decades witnessed the rise of computational approaches to provide quick and fine-grained quantitative linguistic analysis.

For example, Facebook, Instagram, e-commerce websites, and blogs improve customer satisfaction and the overall shopping experience for the customer by allowing customers to rate or comment on the products they have purchased or are planning to purchase3. These visualizations serve as a form of qualitative analysis for the model’s syntactic feature representation in Figure 6. The observable patterns in the embedding spaces provide insights into the model’s capacity to encode syntactic roles, dependencies, and relationships inherent in the linguistic data. For instance, the discernible clusters in the POS embeddings suggest that the model has learned distinct representations for different grammatical categories, which is crucial for tasks reliant on POS tagging. Moreover, the spread and arrangement of points in the dependency embeddings indicate the model’s ability to capture a variety of syntactic dependencies, a key aspect for parsing and related NLP tasks.

For example, a brand could train an algorithm on a set of rules and customer reviews, updating the algorithm until it catches nuances specific to the brand or industry. HyperGlue is a US-based startup that develops an analytics solution to generate insights from unstructured text data. It utilizes natural language processing techniques ChatGPT such as topic clustering, NER, and sentiment reporting. Companies use the startup’s solution to discover anomalies and monitor key trends from customer data. SemEval (Semantic Evaluation) is a renowned NLP workshop where research teams compete scientifically in sentiment analysis, text similarity, and question-answering tasks.

From the learning curve of the GRU model, the gap between the training and the validation accuracy is minimal, but the model at the start begins to underfit. However, when the researcher increases the epoch number, the accuracy increased, which overcomes underfitting. The loss was high with 64% at the first iteration, but it decreases to a minimum in the last epoch to 32%.

It’s designed to house all your valuable data in a convenient, easy-to-digest format. The first step of social media sentiment analysis is to find the conversations people are having about your brand online. Running a social media sentiment analysis program is both an art and a science. Sprout’s sentiment analysis tools provide real-time insights into customer opinions, helping you respond promptly and appropriately. This proactive approach can improve customer satisfaction, loyalty and brand reputation. Finding the right tone on social media can be challenging, but sentiment analysis can guide you.

Here’s an example of positive sentiment from one of Girlfriend Collective’s product pages. Track conversations and social mentions about your brand across social media, such as X, Instagram, Facebook and LinkedIn, even if your brand isn’t directly tagged. Doing so is a great way to capitalize on praise and address criticism quickly. Positive interactions, like acknowledging compliments or thanking customers for their support, can also strengthen your brand’s relationship with its audience. Social sentiment analytics help you pinpoint the right moments to engage, ensuring your interactions are timely and relevant. In the video below, hear examples of how you can use sentiment analysis to fuel business decisions and how to perform it.

VADER calculates the text sentiment and returns the probability of a given input sentence to be positive, negative, or neural. The tool can analyze data from all sorts of social media platforms, such as Twitter and Facebook. Social media sentiment analysis tools can provide valuable insights into how your brand is perceived online. To make your life easier, we’re giving away a free social media sentiment analysis report template.

Moreover, it is also helpful to customers as the technology enhances the overall customer experience at different levels. The datasets generated during and/or analysed during the current study are available from the corresponding author upon reasonable request. \(C\_correct\) represents the count of correctly classified sentences, and \(C\_total\) denotes the total number of sentences analyzed.

On the other end of the spectrum are brands like Patagonia, which is known for its environmental and sustainability efforts. Its social media content regularly focuses on its philanthropic efforts and spotlights global projects aimed at protecting the environment—much to the delight of its customers. Stanley was quick to put out a statement that while they do use lead in their manufacturing process, there is no risk to consumers. While some negative sentiment remains online, trending videos supporting the Stanley brand definitely out number them. By following trends and investigating spikes in positive, negative, or neutral sentiment, you can learn what your audience really wants. This can give you a better idea of what kind of messaging you should post on each social network.

This investigation is of particular significance as it contributes to the development of automatic translation systems. This research contributes to developing a state-of-the-art Arabic sentiment analysis system, creating a new dialectal Arabic sentiment lexicon, and establishing the first Arabic-English parallel corpus. Significantly, this corpus is independently annotated for sentiment by both Arabic and English speakers, thereby adding a valuable resource to the field of sentiment analysis. Our increasingly digital world generates exponential amounts of data as audio, video, and text.

The primary objective of this study is to assess the feasibility of sentiment analysis of translated sentences, thereby providing insights into the potential of utilizing translated text for sentiment analysis and developing a new model for better accuracy. By evaluating the accuracy of sentiment analysis using Acc, we aim to validate hypothesis H that foreign language sentiment analysis is possible through translation to English. Currently, NLP-based solutions struggle when dealing with situations outside of their boundaries. Therefore, AI models need to be retrained for each specific situation that it is unable to solve, which is highly time-consuming. Reinforcement learning enables NLP models to learn behavior that maximizes the possibility of a positive outcome through feedback from the environment. This enables developers and businesses to continuously improve their NLP models’ performance through sequences of reward-based training iterations.