Sentiment Analyzing on Weibo data

Sentiment analyzing is analyzing a sentence to find out that the sentiment of it is positive or negative. Part of the project task of our team is analyzing the sentiment of the comments of 4 domestic mobile phones: Xiaomi M4, Smartisan T1, Huawei Honor6, Meizu MX4. 

Here are the key steps for algorithm：
1.    Read the text and tokenize it.
2.    In every sentence, find the sentiment word, and record its feature (positive or negative) according to the sentiment dictionary and position.
3.    Find the adverb of degree before the sentiment word. When we find one then stop searching. And we will set weights for adverbs of different degrees. And the weights will multiply the sentiment value (assume the primary sentiment value of every sentiment word is 1)
4.    Find all the negation word before the sentiment word. If the number of negation word is odd, then the sentiment value will multiply -1. If the number is even, multiply 1.
5.    If there is ‘!’ in the sentence, every ‘!’ will add 2 sentiment value to the corresponding feature.
6.    Print out positive and negative value and the corresponding percentage of every sentence.
7.    Add all the sentiment value up and print out the positive and negative value and the corresponding percentage of the whole text.
8.    Calculate the average and variance of the positive and negative sentiment for the text.
 
And during the programming, I came up with some problems:
1.       Python is a little troublesome for processing Chinese characters. The encode information should be presented in Unicode.
2.       Python sometimes can’t input the data in txt file completely.
3.       As the Internet words are much different from the standard sentiment dictionary. We should add and edit some words in the sentiment dictionary after we study the linguist habits of the netizen. That will promote the accuracy of the analyzing outcome.
4.       There is some difference of sentiment value between analyzing the whole text and analyzing every sentence and add them together. I think there might be some unnecessary values at the boundary of two sentences. For example, the sentiment word in the beginning of a sentence will look for adverb of degree in the end of the previous sentence.
 
We are  still working on optimizing the outcome. Hope we can achieve our goals.  

Platform Conversion and Regulation

The technology of social media may have already developed to its limit.  More and more functions are added to these platforms, which is more than we need sometimes. But more platforms and more functions don't mean better. We only want to select the best fit platforms which suit our needs at different situations, and find more methods to build it better. It's both the duty of administrators and the users.

For instance, I was used to Baidu Tieba(various forums built by users) and Baidu Knows(collective intelligence by questions and answers provided by users). But now I barely use the two platforms. Because usually I can’t get the information that I need from them. The two platforms are now full of redundant information, which is not good for not only the efficiency of my study but also entertainment. So for study, I will mainly go to Zhihu and CSDN for others’ experience and point of view, because more professional and experienced people are gathering there. I think it is inevitable for platform conversion as our ages and positions vary, and people easily get aesthetic fatigue and tended to find new place to build their regulation when the old is difficult to manage and use.

Also good self-regulation is important for every user. When come more and more users, it will be more likely to be out of control and lead to quarrel, useless information. For instance, Weibo is an open place for users to share their information, and I follow a micro blogger who talk about basic matters of life for other users. At regular intervals, she will delete her followers who have a conflict of values with her and who cause unhappiness in the comments. In this way, she makes the place happy and comfortable for her and the majority of her followers. I think this is a proper way to arrange a little place inside the big social network.  

Sentiment in Social Network

We love comments. The comments under our blogs,  friends' blogs, celebrities blogs, or on the online shopping sites, tourist website... etc. Why? On the one hand, we need information. Is the product worth buying? How to travel a place with less money and find lesser known wonderful attractions? On the other hand, we enjoy finding unknown connections among others, and others' sentiment over a topic. Indeed social network gives us the platform to get nearly all the information we want, but it also leads to information overload.

Unlike normal communication with others, the power of words on the screen sometimes is stronger than we think. Some words full of negative emotion or false information can be spread widely. People who contact those words may get in bad mood, and even eventually form negative thinking patterns, especially teenagers. Also, it can be problems for sentiment analysis research. 

Not all the information is necessary for people, and not all the sentiment matters. In the research, false information should be ignored. But the affection of words on people's mind can't be deleted easily. If we care about the sentiment in social media too much, it may lead to confusing, contradictory and even overwhelming. So we should arrange our time for social network,  learn to distinguish between what's reliable and what isn't. Spending time reading books, thinking independently, and communicating with others directly, are good ways for us, who are living in the information overload world.

The Power of Social Network

In the last century, we could never imagine a day when everyone can be both a role of an information receiver and a deliver at the same time.  And then came the generation of Web 2.0. More and more people become active participators in the construction of network information. It's quite ordinary for us to find what we want to know at any time and any place where it can be connected to the internet, not restricted to the books or meeting with someone to get advice. For instance, if I want to know a product whether it is ' worth buying, my way is to search key words to find comments of the product on Sina Weibo. Soon I can find the price, quality or drawbacks of the product in real time, getting rid of the advertisement from salers or other distracting information. 

Just a few days ago,my friend highly recommended me a bilingual class called Bilingual News on Podcast on iPhone. The content is about contemporary news spoken in terms of Japanese and English. She says she is addicted to the deep gentle voice of the speaker. 

Podcast is also an great achievement of Web 2.0 since it has come out in 2004 with iPodder. Not like traditionally listening to the radio by a small group of professional people, everyone can upload clips of audio or video. And people can subscribe what they like and the files will be updated and downloaded automatically via RSS.  After listened to some periods of this audio podcast, I found the speakers are just ordinary people, instead of professional linguistic teachers.  What attract me most is that after the giving the news, they deliver their opinions about the news freely and make talk interesting. I can get the thoughts of people who living in other countries  with different living experience and culture, even on some sensitive problems, in stead of boring or scaring quarreling and rumors. These won't be written in the books and it can only be achieved by today's social network platform. I believe such way for exchange of thoughts and values will help a lot to the human mutual understanding.

References

Podcast - Wikipedia

http://en.wikipedia.org/wiki/Podcast