Deep learning based opinion mining on youtube

• Main Authors: Solomi, V. Sherlin, Rajana, Swaroop Naidu, Kolluri, Madhan Mohan, Hans, W. Jino
• Format: Conference Proceeding
• Language: English
• Subjects: Algorithms; Artificial intelligence; Customer services; Data analysis; Data mining; Deep learning; Digital cameras; Machine learning; Natural language processing; Product development; Sentiment analysis
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0209061

Opinion mining, sometimes referred to as sentiment analysis, is a method used in natural language processing to determine how the public feels about a certain product. Building a system to gather and organise customer reviews of a product is part of this process, which frequently makes use of machine learning or artificial intelligence to detect emotion in text. This method can be useful for several reasons, including assisting marketers in assessing the success of an advertising campaign or the introduction of a new product, figuring out which features of a product or service are popular among demographics, and spotting any potential weaknesses in a given product. For instance, a website review of a digital camera can be generally favourable but especially critical of its weight. In comparison to more conventional techniques like surveys or focus groups, opinion mining enables vendors to acquire a more thorough insight of consumer sentiment. Using opinion mining to analyse comments on YouTube videos is one example of its use. The automated comprehension, extraction, and processing of textual data from comments enables the recognition of emotive data. In order to assist viewers, save time and choose high-quality material, this research uses the Naive Bayes algorithm and the TextBlob Library to conduct sentiment analysis on YouTube comments, determine the polarity of the remarks, and provide an output in the form of ratings (out of five).