Free Articles, Free Web Content, Reprint Articles
Sunday, July 15, 2018
 
Free Articles, Free Web Content, Reprint ArticlesRegisterAll CategoriesTop AuthorsSubmit Article (Article Submission)ContactSubscribe Free Articles, Free Web Content, Reprint Articles
 

Facebook Data Analysis Can Really Help Deal With Hate Speech

Facebook was among the first social media to appear to the world. Like any other social media, Facebook has had an upsurge in the number of people who signing up. Increase in the number of people leads to an increase in the amount of data that conveyed through the Facebook servers

It also leads to use of informal language which makes it a little bit difficult to perform the data processing. However, it is not completely difficult since there are techniques for performing natural language processing that are being developed.

How natural Language processing can be implemented in Facebook:

Since Facebook has existed for more than a decade now, it has the highest number of users, to the tune of three billion. Therefore, it means if we can do natural language processing Facebook we can be many steps ahead since it could be possible get the meaning of such messages from a very large number of people. This field of natural language processing requires an expert in the field of machine learning together with vast knowledge in machine language translations, word sense disambiguation topic modeling and many others. Machine learning entails coming up with prior data which we refer to as training example. The system should be coded to learn from the data in the training sets and gain experience. It is then from this experience that the system can make decisions when similar data is fed into it.

In this context, Facebook application should be trained using training data which is the prior posts to the Facebook wall. After application of various machine learning algorithms, it is now said to have gained experience. Similar posts to the Facebook wall in future will be easily classified according to the system coding. For instance, a word like ‘stupid’ and ‘polite’ may have been classified as hate speech and ordinary conversation respectively. When the classifier sees the word ‘stupid’ in a piece of text, it classifies it as hate speech.

Implementing PR in social media

Social media analysis for PR is basically analyzing data for the purpose of public relations. In any given organizational set up we need to have a good PR towards the people we deal with be it clients, employers, employees or even fellow colleagues. Therefore it is in order that when we analyze social media data, we should have in mind, the PR of the subject party.

As a matter of fact, sometimes we may carry out data analysis and the results depict a negative image of the concerned party. Therefore we should always keep in mind that PR, brand management and online reputation are very key areas when doing social media analysis, particularly sentiment analysis of social data. We should therefore make sure that we uphold the morals of any given organization when performing data analysis. Furthermore, there should be a way of dealing with customers of an organization in case of any complaint or observation. This is far much possible since most of the people are now active on social mediaComputer Technology Articles, hence they can meet online with automated PR system for social media.

Article Tags: Natural Language Processing, Data Analysis, Hate Speech, Natural Language, Language Processing, Machine Learning, Social Media

Source: Free Articles from ArticlesFactory.com

ABOUT THE AUTHOR


Currently,Natural Language Processing Facebook is not in operation, perhaps due to the privacy terms and conditions that are there in Facebook compared to other social networks like twitter which is entirely public.Social media analysis for PR should also be embraced in a special way possible.



Health
Business
Finance
Travel
Technology
Home Repair
Computers
Marketing
Autos
Family
Entertainment
Law
Education
Communication
Other
Sports
ECommerce
Home Business
Self Help
Internet
Partners


Page loaded in 0.172 seconds