Product Rating through Sentiment Analysis
Abstract : Sentiment analysis is defined as the process of mining of data, view, review or sentence to predict the emotion of the sentence through natural language processing (NLP). The sentiment analysis involve classification of text into three phase “Positive”, “Negative” or “Neutral”.
It analyzes the data and labels the ‘better’ and ‘worse’ sentiment as positive and negative respectively. Thus, in the past years, the World Wide Web (WWW) has become a huge source of raw data generated custom or user. Using social media, e-commerce website, movies reviews such as Facebook, twitter, Amazon, Flipkart etc.
user share their views, feelings in a convenient way. In WWW, where millions of people express their views in their daily interaction, either in the social media or in e-commence which can be their sentiments and opinions about particular thing.
These growing raw data are an extremely high source of information for any kind of decision making process either positive or negative. To analysis of such huge data automatically, the field of sentiment analysis has turn up. The main aim of sentiment analysis is to identifying polarity of the data in the Web and classifying them.
In this 21st century, people are more social in social media, internet, online shopping etc. Thus directly or indirectly online judgments, opinions are eventually gaining great attention. But the real deal is analysis or mining of opinions. Below is the review of some existing solutions available for SA.
These methods are also briefly tabulated in Table.1. OPINE, an unsupervised, web-based information extraction system proposed by Propescu et al. [5] extracted product feature and opinions from reviews.
It identifies product feature, opinion regarding product feature, determines polarity of opinions and then ranks product accordingly [9]. In feature identification, nouns from dataset or reviews are extracted. Frequencies higher than the threshold frequency are kept else discarded.
OPINE’s feature assessor is used to extract explicit features (occurrence of frequent features) [4]. Researchers have used manual extraction rule to extract data [4]. Advancement of OPINE is its domain independency. But fails to find its real life uses as OPINE system is not easily available.
Accuracy Issues
Contextual Misunderstanding: Sentiment analysis algorithms may struggle with context, leading to misinterpretation of sentiments. For example, sarcasm or irony can be challenging to detect.
Ambiguity in Language: Words or phrases with multiple meanings can result in incorrect sentiment classification. For example, “sick” might be positive in a slang context but negative in a medical context.
Language and Cultural Variability
Multiple Languages: Sentiment analysis models trained on one language may not perform well on others. Adapting models to handle different languages and dialects can be complex.
Cultural Differences: Sentiment expressions can vary greatly across cultures, making it hard to generalize sentiment analysis across diverse user bases.
However, those types of online data have several flaws that potentially hinder the process of sentiment analysis. The first flaw is that since people can freely post their own content, the quality of their opinions cannot be guaranteed.
For example, instead of sharing topic-related opinions, online spammers post spam on forums. Some spam are meaningless at all, while others have irrelevant opinions also known as fake opinions [10-12].
The second flaw is that ground truth of such online data is not always available. A ground truth is more like a tag of a certain opinion, indicating whether the opinion is positive, negative, or neutral. The Stanford Sentiment 140 Tweet Corpus [13] is one of the datasets that has ground truth and is also public available.
The corpus contains 1.6 million machine-tagged Twitter messages. Each message is tagged based on the emoticons (as positive, as negative) discovered inside the message.
Real-Time Feedback
Immediate Insights: Sentiment analysis can process customer reviews and feedback quickly, providing near-instantaneous insights into product performance and consumer satisfaction.
Timely Adjustments: Businesses can respond to issues or trends in real time, allowing for faster adjustments and improvements.
Scalability
Handling Large Volumes: Sentiment analysis can efficiently process vast amounts of data, including thousands of reviews or social media posts, which would be impractical to analyze manually.
Automated Analysis: Automation reduces the need for manual labor and speeds up the analysis process.
|