A Study of Sentimental Value Analysis for Tweeting Message

In the recent years, social media analysis has brought several kinds of effective results to us in the domains of industries, education, social science, economics, and so on. This study focuses on Twitter used by students of universities, analyzes tweeting data (messages) by students from Twitter, and calculates sentimental values from the according data. Our approach includes deciding to select and cheese suitable Twitter accounts who tweet interesting messages and acquire a possible mount of their relevant message, specifying and defining value for word of messages for sentimental dictionary, categorizing and listing such a calculated value for each word or expression as item of the dictionary, using these values as "Sentimental Values" for calculation of tweeting messages, calculating Sentimental Values for tweeting message from students of selected universities, checking statistical relation between the above "Sentimental Values" and published data. It also investigates existence of some relations between calculated sentimental values and practical thought of students. At the same time, it tries to discuss whether the above procedure and analysis can visualize conventionally hidden relationship between contents of tweeting messages and characteristic behavior of students in categorized universities.


Introduction
In recent years, young people use social media such as Twitter.And information processing for social media can bring several kinds of effective results to us in the domains of industries, education, social science, economics, and so on.We can expect that our current environment will bring us useful facilities to analyze messages from Twitter based on data mining 1 and their attractive results to visualize normally hidden relations among several kinds of events and phenomena.
This study is to perform some kind of Big Data Analysis 2 as an example of visualization of sentimental words extracted from messages of Twitter and to analyze hidden or unknown relation between sentimental value and the relevant behavior/phenomena.We have defined translation scheme from words in Twitter's message into sentimental values, applied the scheme into data mining with calculation of sentimental values for tweeting messages acquired from Twitter, and demonstrated relation between sentimental values and categorized human behavior.This paper includes the following five sections, namely the next section describes our related works, the third P -587 section illustrates our approach of sentiment analysis for Twitter, the fourth section discusses our results and their visualization of relation between sentimental values and behavior/phenomena, and finally the last section summarizes our conclusion.

Related Work
This section introduces some related works about twitter-based data-mining, knowledge discovery and sentiment analysis.Ley Zhang of HP Laboratory reported in their article 3 "With the booming of microblogs on the Web, people have begun to express their opinions on a wide variety of topics on Twitter and other similar services.Sentiment analysis on entities (e.g., products, organizations, people, etc.) in tweets (posts on Twitter) thus becomes a rapid and effective way of gauging public opinion for business marketing or social studies.However, Twitter's unique characteristics give rise to new problems for current sentiment analysis methods, which originally focused on large opinionated corpora such as product reviews."Mikalai Tsytsarau and Themis Palpanas described in their journal article 4 "In the past years we have witnessed Sentiment Analysis and Opinion Mining becoming increasingly popular topics in Information Retrieval and Web data analysis.With the rapid growth of the user-generated content represented in blogs, wikis and Web forums, such an analysis became a useful tool for mining the Web, since it allowed us to capture sentiments and opinions at a large scale.Opinion retrieval has established itself as an important part of search engines.Ratings, opinion trends and representative opinions enrich the search experience of users when combined with traditional document retrieval, by revealing more insights about a subject.Opinion aggregation over product reviews can be very useful for product marketing and positioning, exposing the customers' attitude towards a product and its features along different dimensions, such as time, geographical location, and experience."Alexandra Balahur of European Commission Joint Research Centre pointed out in her international conference paper 5 "This paper presents a method for sentiment analysis specifically designed to work with Twitter data (tweets).The main contributions of this work are: a) the pre-processing of tweets to normalize the language and generalize the vocabulary employed to express sentiment; b) the use minimal linguistic processing, which makes the approach easily portable to other languages; c) the inclusion of higher order ngrams to spot modifications in the polarity of the sentiment expressed; d) the use of simple heuristics to select features to be employed; e) the application of supervised learning using a simple Support Vector Machines linear classifier on a set of realistic data."Referring the above works and our previous research, we have tried to apply sentiment analysis and visualize normally hidden relation between sentimental values and behavior/phenomena.

Sentimental Analysis as Data Mining
Schematic diagram of our sentimental analysis flow can be illustrated and explained in Fig. 1.We have focused on specific data based on speakers' attributes as target users.After connecting Twitter as Web-based Social Media, our system has acquired an amount of data to be analyzed with "twpro API" and "Twitter API".With these data, the main part of our system, which is constructed with multiple numbers (#1-#n) of modules, performs a series of manipulations for information processing from data into generated results which include some temporary files.From top to bottom in Fig. 2, our system sequentially defines target users of Twitter, obtains tweeting messages from defined users, prepares Sentimental dictionary based on those messages, computes sentimental values for each users, and then generates sets of such values focused groups (universities in our case).This part is the main body of our system, showing "System Configuration and Processing Flow" and can P -588 accomplish some kind of sentiment analysis and generate our interesting results.We have obtained related relation between the above files based on messages from Twitter.With the sentimental value for each university, we will compare the published data and try to visualize normally hidden relation between tweeting messages and behavior and/or phenomena in the next section.

Results and Discussion
This section compares results in 2014 and 2015 and then discusses comparison of sentimental values in 2014 and 2015.

Comparison of analyzed results in 2014 and 2015
For

Discussion of Sentimental Values in 2014 and 2015
By comparison with results in 2014 and 2015, we discuss our approach of sentimental analysis.Table 1 shows our results of sentimental analysis for tweeting messages from students of previously defined universities, whose identifier, IDs are specified from 001 to 026 in Table

Conclusion
We have tried to demonstrate and visualize normally hidden or unknown relations between messages from Twitter and information of published documents by means of sentimental analysis.For example, we have been analyzing relationship between sentimental analyzed results and several information from the above documents, such as population of district, number of governmental universities' students, number of Starbucks coffee shops, and so on.

Fig. 1 .
Fig.1.Schematic Diagram for System Configuration of Data Mining of Tweeting Message and Sentimental Analysis.

Fig. 2 .
Fig.2.System Flow of Data Mining of Tweeting Message and Sentimental Analysis with Dictionary of Sentiment Value.

Fig. 3 .
Fig.3.Relation between Sentimental values and numbers of Starbucks Coffee shops.

Fig. 4 .
Fig.4.Relation between Sentimental values and numbers of Starbucks Coffee shops.

Table 1 .
.In 2014, there is only one, ID=018, P -589 whose score of sentimental values is more than 0.5, namely others are less than 0.5, so almost in 2014 have tweeted negative messages.In 2015, sentimental values have varied from 0.4 to 0.90 approximately.Fig.5.shows bar chart about normalized sentimental values in 2014 and 2015.Generally speaking, such values in 2014 are relatively lower than ones in 2015.We can assume that two patterns of relations between 2014 and 2015.One pattern is a group of students who have tweeted negatively in 2014 and 2015.Another includes students who have tweeted negatively in 2014 but done positively in 2015.Fig.6 demonstrates that targets can be divided into two groups described before.Sentimental Values in 2014 and 2015.