1.1 as products, services, organizations, individuals, issues, events,

1.

1  IntroductionSentiment analysis is the field of study that analyzespeople’s opinions, sentiments, evaluations, appraisals, attitudes, and emotionstowards entities such as products, services, organizations, individuals,issues, events, topics, and their attributes. Sentiment analysis, which is alsocalled opinion mining, involves in building a system to collect and examineopinions about the product made in blog posts, comments, reviews or tweets. Themeaning of opinion itself is still very broad. Sentiment analysis and opinionmining mainly focuses on opinions which express or imply positive or negativesentimentsNow there is a huge volume of opinionated data in thesocial media on the Web.

Without this data, a lot of sentiment analysis wouldnot have been possible. Not surprisingly, the inception and the rapid growth ofsentiment analysis coincide with those of the social media. In fact, sentimentanalysis is now right at the center of the social media analysis. Hencesentiment analysis not only has an important impact on NLP, but may also have aprofound impact on management sciences, political science, economics, andsocial sciences as they are all affected by people’s opinions.Opinions arecentral to almost all human activities because they are key influencers of ourbehaviors.

Whenever we need to make a decision, we want to know others’opinions. In the real world, businesses and organizations always want to findconsumer or public opinions about their products and services. Individualconsumers also want to know the opinions of existing users of a product beforepurchasing it, and others’ opinions about political candidates before making avoting decision in a political election. In the past, when an individual neededopinions, he/she asked friends and family. When an organization or a businessneeded public or consumer opinions, it conducted surveys, opinion polls, andfocus groups.

Best services for writing your paper according to Trustpilot

Premium Partner
From $18.00 per page
4,8 / 5
4,80
Writers Experience
4,80
Delivery
4,90
Support
4,70
Price
Recommended Service
From $13.90 per page
4,6 / 5
4,70
Writers Experience
4,70
Delivery
4,60
Support
4,60
Price
From $20.00 per page
4,5 / 5
4,80
Writers Experience
4,50
Delivery
4,40
Support
4,10
Price
* All Partners were chosen among 50+ writing services by our Customer Satisfaction Team

Acquiring public and consumer opinions has long been a hugebusiness itself for marketing, public relations, and political campaigncompanies.With theexplosive growth of social media (e.g., reviews, forum discussions, blogs,micro-blogs, Twitter, comments, and postings in social network sites) on theWeb, individuals and organizations are increasingly using the content in thesemedia for decision making.

Nowadays, if one wants to buy a consumer product,one is no longer limited to asking one’s friends and family for opinionsbecause there are many user reviews and discussions in public forums on the Webabout the product. For an organization, it may no longer be necessary toconduct surveys, opinion polls, and focus groups in order to gather publicopinions because there is an abundance of such information publicly available. In recentyears, it was witnessed that opinionated postings in social media have helpedreshape businesses, and sway public sentiments and emotions, which haveprofoundly impacted on our social and political systems. Such postings havealso mobilized masses for political changes such as those happened in somecountries.

It has thus become a necessity to collect and study opinions on theWeb. Of course, opinionated documents not only exist on the Web (calledexternal data), many organizations also have their internal data, e.g.

,customer feedback collected from emails and call centers or results fromsurveys conducted by the organizations.Due to theseanalysis work, industrial activities have flourished in recent years. Sentimentanalysis applications have spread to almost every possible domain, from consumerproducts, services, healthcare, and financial services to social events andpolitical elections. A key feature of social media is that it enables anyonefrom anywhere in the world to freely express his/her views and opinions withoutdisclosing his/her true identify and without the fear of undesirableconsequences. These opinions are thus highly valuable. Twitter is a place wheremany people express their views in the form of tweets.

In general, sentiment analysis has been investigatedmainly at three levels. In document level the main task is to classify whethera whole opinion document expresses a positive or negative sentiment. This levelof analysis assumes that each document expresses opinions on a single entity.In sentence level the main task is to check whether each sentence expressed apositive, negative, or neutral opinion. This level of analysis is closelyrelated to subjectivity classification, which distinguishes objective sentencesthat express factual information from subjective sentences that expresssubjective views and opinion. Document level and the sentence level analyses donot discover what exactly people liked and did not like.

Aspect level performsfiner-grained analysis. Instead of looking at language constructs (documents,paragraphs, sentences, clauses or phrases), aspect level directly looks at theopinion itself. Sentiment analysis played a great role in the area ofresearches done by many, there are many methods to carry out sentimentanalysis.

Still many analyses are going on to find out better alternatives dueto its importance in this scenario. Some of the methods are Machine learning strategies.They work by training an algorithm with a training data set before applying itto the actual data set. Machine learning techniques first trains the algorithmwith some particular inputs with known outputs so that later it can work with newunknown data. Some of the most renowned works based on machine learning is Support Vector Machine.

It is a non-probabilisticclassifier in which a large amount of training set is required. It is done byclassifying points using a (d-1) dimensional hyper plane. SVM finds a hyperplane with largest possible margin. Support Vector Machines make use of theconcept of decision planes that define decision boundaries. A decision plane isone that separates between a set of objects having different class membership. Naïve Bayes Method is alsoa most renowned approach in machine learning strategies. Itis a probabilistic classifier and is mainly used when the size of the trainingset is less. In machine learning it isin family of sample probabilistic classifier based on Bayes theorem.

Theconditional probability that an event Xoccurs given the evidence Y is determined by Bayes rule.   A Maximum Entropy (ME) classifier, or conditionalexponential classifier, also belongs to machine learning strategies, it is parameterizedby a set of weights that are used to combine the joint-features that aregenerated from a set of features by an encoding. The encoding maps each pair offeature set and label to a vector. ME classifiers belong to the set ofclassifiers known as the exponential or log-linear classifiers, because theywork by extracting some set of features from the input, combining them linearlyand then using this sum as exponent.

Lexicon Based strategies work on an assumption thatthe collective polarity of a sentence or documents is the sum of polarities ofthe individual phrases or words. In this strategy we have two most ownedapproaches, they are corpus based and dictionary based. The dictionary method isgoverned by the use of a dictionary consisting pre-tagged lexicons. The inputtext is converted to tokens by the Tokenizer. Every new token encountered isthen matched for the lexicon in the dictionary. If there is a positive match,the score is added to the total pool of score for the input text. For instance,if “dramatic” is a positive match in the dictionary then the total score of thetext is incremented.

Other-wise the score is decremented or the word is taggedas negative. Though this technique appears to be amateur in nature, itsvariants have proved to be worthy.The computational speed and efficiency ofdictionary-based approaches to sentiment analysis, together with theirintuitive appeal, make such approaches an attractive alternative for extractingemotional context text. At the same time in both dictionariesbased and corpus based approaches re-constructed dictionaries for use withmodern standard U.S. English have the advantage of being exceptionally easy touse and extensively validated, making them strong contenders for applicationswhere the emotional content of the language under study is expressed in conventionalways.

1.1  IntroductionSentiment analysis is the field of study that analyzespeople’s opinions, sentiments, evaluations, appraisals, attitudes, and emotionstowards entities such as products, services, organizations, individuals,issues, events, topics, and their attributes. Sentiment analysis, which is alsocalled opinion mining, involves in building a system to collect and examineopinions about the product made in blog posts, comments, reviews or tweets. Themeaning of opinion itself is still very broad. Sentiment analysis and opinionmining mainly focuses on opinions which express or imply positive or negativesentimentsNow there is a huge volume of opinionated data in thesocial media on the Web. Without this data, a lot of sentiment analysis wouldnot have been possible. Not surprisingly, the inception and the rapid growth ofsentiment analysis coincide with those of the social media.

In fact, sentimentanalysis is now right at the center of the social media analysis. Hencesentiment analysis not only has an important impact on NLP, but may also have aprofound impact on management sciences, political science, economics, andsocial sciences as they are all affected by people’s opinions.Opinions arecentral to almost all human activities because they are key influencers of ourbehaviors. Whenever we need to make a decision, we want to know others’opinions.

In the real world, businesses and organizations always want to findconsumer or public opinions about their products and services. Individualconsumers also want to know the opinions of existing users of a product beforepurchasing it, and others’ opinions about political candidates before making avoting decision in a political election. In the past, when an individual neededopinions, he/she asked friends and family. When an organization or a businessneeded public or consumer opinions, it conducted surveys, opinion polls, andfocus groups. Acquiring public and consumer opinions has long been a hugebusiness itself for marketing, public relations, and political campaigncompanies.

With theexplosive growth of social media (e.g., reviews, forum discussions, blogs,micro-blogs, Twitter, comments, and postings in social network sites) on theWeb, individuals and organizations are increasingly using the content in thesemedia for decision making. Nowadays, if one wants to buy a consumer product,one is no longer limited to asking one’s friends and family for opinionsbecause there are many user reviews and discussions in public forums on the Webabout the product. For an organization, it may no longer be necessary toconduct surveys, opinion polls, and focus groups in order to gather publicopinions because there is an abundance of such information publicly available. In recentyears, it was witnessed that opinionated postings in social media have helpedreshape businesses, and sway public sentiments and emotions, which haveprofoundly impacted on our social and political systems. Such postings havealso mobilized masses for political changes such as those happened in somecountries. It has thus become a necessity to collect and study opinions on theWeb.

Of course, opinionated documents not only exist on the Web (calledexternal data), many organizations also have their internal data, e.g.,customer feedback collected from emails and call centers or results fromsurveys conducted by the organizations.

Due to theseanalysis work, industrial activities have flourished in recent years. Sentimentanalysis applications have spread to almost every possible domain, from consumerproducts, services, healthcare, and financial services to social events andpolitical elections. A key feature of social media is that it enables anyonefrom anywhere in the world to freely express his/her views and opinions withoutdisclosing his/her true identify and without the fear of undesirableconsequences. These opinions are thus highly valuable. Twitter is a place wheremany people express their views in the form of tweets. In general, sentiment analysis has been investigatedmainly at three levels. In document level the main task is to classify whethera whole opinion document expresses a positive or negative sentiment.

This levelof analysis assumes that each document expresses opinions on a single entity.In sentence level the main task is to check whether each sentence expressed apositive, negative, or neutral opinion. This level of analysis is closelyrelated to subjectivity classification, which distinguishes objective sentencesthat express factual information from subjective sentences that expresssubjective views and opinion. Document level and the sentence level analyses donot discover what exactly people liked and did not like. Aspect level performsfiner-grained analysis. Instead of looking at language constructs (documents,paragraphs, sentences, clauses or phrases), aspect level directly looks at theopinion itself.

Sentiment analysis played a great role in the area ofresearches done by many, there are many methods to carry out sentimentanalysis. Still many analyses are going on to find out better alternatives dueto its importance in this scenario. Some of the methods are Machine learning strategies.They work by training an algorithm with a training data set before applying itto the actual data set.

Machine learning techniques first trains the algorithmwith some particular inputs with known outputs so that later it can work with newunknown data. Some of the most renowned works based on machine learning is Support Vector Machine. It is a non-probabilisticclassifier in which a large amount of training set is required. It is done byclassifying points using a (d-1) dimensional hyper plane.

SVM finds a hyperplane with largest possible margin. Support Vector Machines make use of theconcept of decision planes that define decision boundaries. A decision plane isone that separates between a set of objects having different class membership. Naïve Bayes Method is alsoa most renowned approach in machine learning strategies. Itis a probabilistic classifier and is mainly used when the size of the trainingset is less. In machine learning it isin family of sample probabilistic classifier based on Bayes theorem. Theconditional probability that an event Xoccurs given the evidence Y is determined by Bayes rule.

  A Maximum Entropy (ME) classifier, or conditionalexponential classifier, also belongs to machine learning strategies, it is parameterizedby a set of weights that are used to combine the joint-features that aregenerated from a set of features by an encoding. The encoding maps each pair offeature set and label to a vector. ME classifiers belong to the set ofclassifiers known as the exponential or log-linear classifiers, because theywork by extracting some set of features from the input, combining them linearlyand then using this sum as exponent. Lexicon Based strategies work on an assumption thatthe collective polarity of a sentence or documents is the sum of polarities ofthe individual phrases or words. In this strategy we have two most ownedapproaches, they are corpus based and dictionary based. The dictionary method isgoverned by the use of a dictionary consisting pre-tagged lexicons. The inputtext is converted to tokens by the Tokenizer. Every new token encountered isthen matched for the lexicon in the dictionary.

If there is a positive match,the score is added to the total pool of score for the input text. For instance,if “dramatic” is a positive match in the dictionary then the total score of thetext is incremented. Other-wise the score is decremented or the word is taggedas negative. Though this technique appears to be amateur in nature, itsvariants have proved to be worthy.The computational speed and efficiency ofdictionary-based approaches to sentiment analysis, together with theirintuitive appeal, make such approaches an attractive alternative for extractingemotional context text. At the same time in both dictionariesbased and corpus based approaches re-constructed dictionaries for use withmodern standard U.

S. English have the advantage of being exceptionally easy touse and extensively validated, making them strong contenders for applicationswhere the emotional content of the language under study is expressed in conventionalways.