Data Mining

Abstract- Social Network Analysis (SNA) is used to research the vector relations, connections and human aptitudes. We presented a vector space model (VSM) [1] to extract the human aptitudes out of text corpus of a particular user overall present in a social network which helps in personality judgment. We create a list or dictionary of all emotionally polluted words present in English language. It is concluded that the application of vector space model can obtain the hidden human thinking and reaction to different situations in different times with different persons by text mapping. It also helps to categorize and rank the users on social network into specific categories like extremism, moderate, hatred, patriotism etc.

People in the world are divided into different races by geographical conditions and cultures, their aptitudes and attitudes are different in different situations depict their tendency or inclination to the specific person or situation in different times. If we talk about human psyche every human in this earth have different emotions like hatred, love, religious mindset, extremism, homosexuality, etc. The main thinking of human person is dominated by one of the said emotional aptitudes. It is very difficult to judge the human psyche but it actually depicts in the communication. Internet is a source of communication and it provides number of platform for the communication among masses of different religions and races to share information among each other. Social Networks like Facebook and different forums provide us a rich emotionally environment where people share thoughts and connection between each other. On the social forums like Facebook people tag posts and get tagged in many pictures, quotes and other material where a number of users post their views and comments. So in a view that human communication depicts its nature and way of thinking, so the overall dataset of a particular user on a social networks, it can be used as an aptitude test and one's personality can be judged by it.

It is important in NLP to find the similarity between two words on the basis of text to text similarity using lexical analysis. This paper intends to utilize classical VSM to implement personality judgment from dataset extracted from social network.

A dictionary would be created comprised of all emotional words in English language and all the tokens are mapped or matched with the dictionary containing emotionally polluted words. The frequency of all matched words in a specific category like extremism, hatred, affection etc would show one's personal inclination, also the greater the frequency of matched words in a social network also rate the user globally in the specific category.