What is social media analytics (update)

1.Background – why we do social media analytics

Nowadays, social media such as Weibo, Facebook, twitter and Youtube have penetrated into all aspects of our lives. Users can interact with friends or strangers in various social media, from text chat to video sharing, from topic interaction to precision marketing. Social media has created a barrier-free and two-way exchange of information for billions of people. Therefore, it is more and more important to analyze and understand the complex information flow and its influence in social media.

2.Concept – what is social media analytics

First, social media is a platform based on Internet technology for generating and exchanging user related information. Therefore, social media analytics is to collect and sort out all kinds of data on the platform through certain technical means, refine and visualize the collected data through quantitative or qualitative analysis methods, so as to form valuable information, such as user behavior, user preference, etc.,which can help us achieve relevant goals, such as precision advertising and commercial promotion.

3.Features – differences from traditional media platforms:

In general, social media such as Facebook and twitter have the following characteristics:

①User centered and interactive media design can continuously improve user experience.

② Originality and simplicity: for example, the number of words in Weibo is limited to 140, which promotes the exponential growth of original content.

③ Convenience: information sharing is convenient and efficient, information can be  released anytime and anywhere, and the publishing speed is faster than traditional media, which is suitable for real-time data analysis.

④ Information acquisition has a strong selectivity: users match their interests with other users and filter out the content they don’t like.

⑤ Users are both producers and disseminators of content

4. Steps of social media analytics:

4.1 crystallize purpose

Each person who conducts social media analytics should first crystallize purpose of their own analytics, such as business decision-making, precise advertising, reducing customer service costs, competitive product analysis or obtaining feedback of related products. Because of different purposes, data sources as well as data collection methods and analysis methods are often different.

4.2 data mining

①What kind of data is worth mining?

In social media analytics, not all data is worth mining, not only the data with authenticity is worth analyzing. Most of the time, as long as the data that can affect social media users are valuable data.

In 2018, through analyzing social robots on Twitter^[1], Zakaria el Hjouji of MIT and other scholars found that the public opinion of social networks can be significantly influenced by a small number of social robots, even though the information they generate is usually easy to identify. This sounds incredible, but it’s often used in activities like marketing and public relations. Therefore, in social media analytics, as long as the data on the platform has a certain influence on users and can contain the value output of users, subjective emotions, feelings, etc., we can mine and use these data.  

②how to mine data

Generally, data mining has the following methods: classification, regression, clustering, web page mining, association rules and deviation analysis. We can also use some auxiliary tools of social media analytics, for example, if you want to collect and analyze the detailed data of user activities on Facebook or twitter, you can choose Hootsuite, which can establish detailed user behavior reports for you, automatically filter relevant factors and obtain real-time feedback from users; if your research scope involves many social media, then social Mention is a good choice. It covers hundreds of mainstream social media, including twitter, FriendFeed, youtube, etc. it can help you integrate the content on the platform into a unified information flow, so that analysts can easily understand the hot topics on the mainstream media in real time. Generally speaking, the process of mining is to use text mining, natural language processing and other technologies, with the help of auxiliary tools, to transform qualitative data into quantitative data, and explore the information hidden in it.

5. Example：

Next, I will give you an example about social media ananlytics. In this example, you will start to operate spark streaming to conduct social media analytics, so as to record the country or region where the real-time updated tweets on twitter are located in real time. People who are interested in it can click the following links:

https://www.ibm.com/developerworks/cn/analytics/blog/analyze-social-media-data-real-time/index.html

Reference

[1]Zakaria el Hjouji.The lmpact of Bots on Opinions in Social Networks[DB\OL].CoRR abs/1810.12398(2018)