User Analytics in Online Social Networks: Evolving from Social Instances to Social Individuals
, Stylianos Georgilas 1, Giannis Haralabopoulos 2 and Ioannis Anagnostopoulos 1
Round 1
Reviewer 1 Report
This paper presents a framework that extracts insights about social network users across multiple platforms. Their idea is to study individual behavior in different platforms and they conclude that individuals do not behave the same way across platforms.
The work is interesting. The methodology appears sound. The system design choices are explained. The evaluation methodology is sound. The influence metrics used are appropriate.
The only issue with the paper is with the presentation. There are several grammatical and spelling errors. In addition, the tables in the section 4 do not highlight the information very clearly. Perhaps a more visual representation is better.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The paper proposes a framework to analyze online social network users’ behavioral and disseminated content patterns. In the case of users maintaining accounts in multiple OSNs, these are not regarded as distinct accounts, but rather as the same individual with multiple social instances. The analyzed dataset includes data about 57 social individuals with their accounts on Facebook, Instagram, and Twitter.
The paper is structured in 7 sections, describing related works, the proposed approach for OSNs analytics, the definition of influential metrics in OSNs, experimental results, conclusions, and further works. It includes 31 figures and 8 tables. The list of references has 25 positions, including 3 self-citations.
If we look at the description of the architecture presented in subsection 3.1, in my view the Processing layer is missing a function solving user's requests that imply access to the Persistence storage.
In the first paragraph of section 6, the first definition of social conversation (in line 642) should be deleted because is similar to the social acceptance definition.
Regarding the presentation of experimental results, section 5 is unusually long (13 pages!). I propose the authors split it into two subsections with the first one dedicated to the presentation of the experimentation methodology (objectives, steps, media objects considered and their expected relevance, and so on), and the second one presenting and assessing the experimental results, according to the methodology steps. If a reader doesn't have enough time or interest to read this subsection, she/he will still have the possibility to analyze the synthesis of the experimental findings in the Conclusions section.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
