Review of Sentimental Analysis Methods using Lexicon Based Approach (Meta,…
Review of Sentimental Analysis Methods using Lexicon Based Approach
Computational analysis of public emotions and attitude towards a particular subject
Immensely useful in social media monitoring as it allows us to gain an insight of the public opinion behind certain topics
There are number of articles presented every year in the SA fields. The number of articles in this has increased manifold.
This creates a need to have survey papers that summarize the recent research trends and directions of SA.
The pioneering work of figuring out application and challenges in the field of SA was presented by Pang and Lee  and Liu . They mentioned the techniques used to solve each problem in SA.
SENTIMENTAL ANALYSIS PROCESS
Removing Non English words
Removing Uniform Resource Locators
Slang word translation
Removing extra letters from words
Pre-processing the data reduces the noise in which helps to improve the performance of the classifier. Pre-processing also speeds up the classification process, thus helping in real time SA.
Pre-processing is the process of cleaning the data readying the text for classification. Online texts contain usually lots of noise and unnecessary parts such as tags, scripts.
We aim to find out the opinionative words or phrases that best describes the context which we are dealing with
Sentiment Classification (SC) techniques can be divided into parts namely, ML approach and lexicon based
Based on training an algorithm
Mostly classification on a set of selected features for a specific mission and then test on another set whether it is able to detect the right features and give the right classification
The ML approach can be further divided into parts namely supervised and unsupervised learning methods. The supervised learning methods make use of a large number of labeled training documents.
In the case where it is difficult to find the labeled training documents, the unsupervised methods are used.
Sentiment extraction involves spotting sentiment words within a particular sentence. This is typically achieved using a dictionary of sentiment terms and their semantic orientations.
Dictionary-based approach has some disadvantages associated with them. For example, the sentiment word „low‟ in the context of “calories” might have a positive polarity, whereas “low” in the context of “video resolution” is of negative polarity.
The lexicon-based approach involves calculating orientation for a document from the semantic orientation of
words or phrases in the document
Feature selection is mostly integrated in Machine Learning (ML) algorithms like SVM, Neural Networks (NN), k-Nearest Neighbours (KNN), etc. as the very first step.
Feature selection will also reduce the over-fitting of the learning
scheme to the training data. During this process, it is also important to find a good trade-off between the
richness of features and the computational constraints involved when solving the categorization task.
Arun Kumar Solanki