Sentiment Analysis Model

2297 Words10 Pages

Rule based Sentiment Analysis to explore Multiple Features for online product reviews
A.M.Abirami
Asst. Professor, Dept. of IT Thiagarajar College of Engineering Madurai, Tamil Nadu, India abiramiam@tce.edu Dr.A.Askarunisa
Head, Dept. of CSE Vickram College of Engineering Enathi, Tamil Nadu, India nishanazer@yahoo.com J.Santhana Preethi
G.Sathya Priya
Dept. of Information Technology Thiagarajar College of Engineering, Madurai, Tamil Nadu, India santhanapreethi28@gmail.com sathyapriyagurusamy@gmail.com
ABSTRACT
Sentiment analysis is the emerging technique by which one can extract information from vast amount of data dispersed in web documents to improve knowledge management, business intelligence, marketing, sales, etc. This can be used to determine …show more content…

Karan Chawla et.al [2] proposed a method for sentiment analysis on tweets having conjunctions, connectives, modals and conditionals using discourse relations for polarity detection of tweets. Xiaowen Ding et. al [3] built Opinion Observer, a linguistic rule based system, which was able to detect the polarity of opinions context based and with the use of opinion aggregation function. Damien Poirier et.al [4] developed a model for opinion analysis on movie reviews using machine learning techniques and NLP techniques by building the dictionaries. Guang Qiu [5] et. al proposed a method to extract new sentiment words from sentences and assigned polarity for them using double propagation …show more content…

The sentences where the use of the conjunction or the presence of a double negative has a direct impact on the overall sentiment of the review are identified.
3.2 POS Tagging
In this step, the sentences in the data set collection are tokenized using the POS tagger of Stanford [15]. During this process, a part of speech such as noun, verb, adverb, adjective, conjunctions, negations and the like are assigned to every word in the sentences. It has been made sure that the conjunctions or negatives present in the sentences are tagged correctly using General Inquirer’s word dictionary [14].
3.3 Sentiment Detection
Sentiments are detected for each word using General Inquirer as positive, negative, strong, weak, pleasure, pain and feel.
A conjunction is a word that is used to link words, phrases, or clauses. It is often used to express the relationship between two clauses or phrases in a sentence. A double negative always renders a positive effect i.e. not dislike means like. Conjunctions and double negations have the capability to completely invert the polarity or sentiment of the

Open Document