Social networks consist of entities and links between them. Entities are human and organizations, links between them are social connections such as friendship and family ties[1], in Figure 1, one human is the entity and the link connect he and other entities. Social media is a representative way to implement interaction of users[2], the relationship of each online identities on social media forming a network of social ties. This kind of connection contains a lot of information because of rich services provided by social media, which result in a large amount of data can be used for certain purpose. This is the origin of social media analytics in my understanding.
Definition
The more complete definition of social media analytics is to utilize the data collected in social media platform, with specific objective, analysis from the perspective of content and networking, finally contributes to informed and insightful decision making or prediction[3]. It can be found from the definition that the process of social media analytics is roughly divided into five parts, which includes motivation and purpose, data collection, analysis, interpretation, prediction and decision-making[4]. Figure 2 shows the general process of social media analytics.
The purpose of social media analytics may vary from case to case but can be summarized and classification into two categories, one is to study social behavior of online identities (e.g.， interaction of social media users, newly emerging topics, etc.), the other is to make improvement on online service in order to better support social activities (e.g.， improvement version of recommendation system )[5].
To facilitate the analysis of social media, collecting user-based data on social media platforms, which are users’ personal information, demographic information and data generated by users’ online behavior [6]. With this range of data, mathematics and computational methods are applied to representation information from both humanities and technologies aspects[7].
Data source
Source data are collected from social media such as Twitter, Facebook and Instagram. The first step is to scrape data, those data is accessible when applying APIs. The raw data is unstructured and contains many noises, the next step is to clean the raw data, normalize the text and remove unstructured entries. After then the data can be used for analysis[8].
Analysis methods
Three methods are introduced for analyzing.
Text analysis
Text analysis is a technique used for content analysis. In social media spectacle, many commit and review are made to express users’ point of view. because of the massive text information in social media, quantitative methods are required to classify the data. Text classification, such as one important analysis technique in social media analysis, sentiment analysis, is a considerable method which can be classified as supervised and unsupervised leaning. With this method data can be detected positive/negative (supervised) or cluster into groups(unsupervised).
Social network analysis
Network analysis study the relationship between online identities. From the relationship in between online identities, some communities who always share similar opinions can be found as well as influence factor of each identity. With the analysis result there are possibilities to influence opinions in social networks.
Trend analysis
Hot topics and newly emerging trends are keeping change on social media, trends analysis can be helpful to predict the coming one. To achieve this, a model or framework should be built to represent the trend and life cycle of one topic. One method is suggested to identify items that attract lots of attention in early stage. Social media activities are modeled with a stochastic model that intuitively captures the concept of attention gathering information items [9].
References
[1] Lecture notes page 15.
[2] Lecture notes page 20
[3] https://en.wikipedia.org/wiki/Social_media_analytics
[4] Lecture notes page 28
[5] Lecture notes page 25
[6] M. Mathioudakis, in Encyclopedia of Database Systems, edited by L. Liu and M. T. Özsu (Springer New York, New York, NY, 2018), pp. 3528-3532.
[7] Lecture notes page 26
[8] Batrinca, B. & Treleaven, P.C. AI & Soc (2015) 30: 89.
[9] Stieglitz, S., Dang-Xuan, L., Bruns, A. et al. Bus Inf Syst Eng (2014) 6: 89.
IEMS5720-Social Networking-Xiong Wy 
Small bear! I am Luman! I like your layout design of your article, which is very concise and structured, which is good for me to read! Besides, one of your topic “Analysis methods” reminds me of the MCL assignments, which is related to SMA.😋
LikeLike
MCL is important in SMA, many methods are introduced for text analysis. Data collected from social media contains many comments, reviews and many text descriptions so that MCL are widly used in SMA.
LikeLike
The first paragraph tells us why social media analytics exists: There are large amout of data which can be used for certain purpose after analysis.
Then this article demonstrates the pipeline of social media analysis. Besides, it stresses the data source and analysis methods for the reason for their status in pipeline.
BTW, I think you should change the title “Definition” to “Pipeline” or other related things because you have already put the definition in the first sentence. The rest are work flow of social media analytics, which “definition” is not precise enough to generalize.
LikeLike
Sorry for typo: “Besides, it stresses the data source and analysis methods for the reason for their status in pipeline” should be “Besides, it stresses the data source and analysis methods for their status in pipeline”.
LikeLike
Thanks for your advice! Pipeline is more accurate.
LikeLike
Your post is logic and thorough with detailed definition and examples.
Today, with rapid development of Internet, we’ve already obtained a lot of tools to get enough raw data (sometimes programming skills even are not required), but how to deal with and use these raw data is becoming a hot topic.
As you mentioned, we have several analysis methods, text, social network and trend analysis are applied to help us to get a clear logic line in this chaotic Internet.
I totally love this article and I think even some readers with other backgrounds can also understand. Hope more specific examples can be provided in your next blog~!
LikeLiked by 1 person
So glad you like this blog. I agree that data processing is a key point in social media analytics, when people talk about SMA methods, what they mean is the way to analyze social media data. For example, Technique methods such as NLP and quantitative analysis apply statistics are widely used in practice.
LikeLike
In the part that is before “Definition”, you introduce two terms include social networking and social media, which are used to pave the way for the following definition of SMA. This gradual process makes your definition of SMA easier to understand. However, in the part of “Definition”, you give more information about the workflow of SMA. It would be better if you illustrate each procedure in the individual subtitle. In another part named “Analysis methods”, “Text analysis” is one of the SMA methods, but the rest two, actually, are specific SMA tasks (scenarios) rather than methods. I am trying to give you some useful suggestions, and hope you don’t mind. Come on, babe!
LikeLike
First of all, thanks for your advice, but I don’t think the rest two parts are SMA tasks, the same as “Text analysis”, they are methods. I think you might misunderstand “method”, text analysis does not mean a specific tool(e.g., NLP) but analysis object. Thanks anyway for reading.
LikeLike
Hi! 🐻, I’m coming! ~ ~ ~
After reading your blog, I can feel your article structure is clear and rich sense of hierarchy, from definition to data source finally analyzed three methods. I think you have summarized the purpose of social media analysis very accurately. Indeed, We want to study social behavior of online identities or make improvement on online service in order to better support social activities. Looking forward to your next blog and also miss you so much!! Btw, your home screen is so cute, I love it!
LikeLike
Thanks for reading! What I have done is to summarize the definition and purposes from what I read, glad you like it.
LikeLike
Your blog clearly tells me what is social media and the flow chart makes the social media analytic process more easier to understand. There are many information hiding in the social media, but finding and filtering the valuable part from the vast amount of data could be a challenging thing. I still have a lot to learn.
LikeLike