SocioScope: A framework for understanding Internet of Social Knowledge

Abstract

Social data is one of important material for representing social activities. However, efficiently collecting social data from multiple sources is extremely hard due its big-data attributes. Moreover, discovering hidden patterns between social data to understand our society is actually a big challenge. Especially, knowledge around us is not stable but dynamically change over time, therefore, we need to refresh it promptly. SocioScope is created as an automatically tool to overcome aforementioned problems. In addition, other researchers can utilize SocioScope as a framework for reducing their effort time with essential tasks (i.e., collecting data, pre-processing data, and analyzing data).

Introduction

Due to the growth of Internet of Things (IoT), every devices (e.g., wireless sensor networks, GPS, control systems) are connected. Data is generated through every activities and it brings human into era of big data. From a statistics [1] in 2013, social data increased quickly in every sides (i.e., volume, velocity, variety) and reached 4.4 zettabytes. By mining hidden knowledge from this recourse, we can achieve huge benefits for our society. However, social data is still raw and exists without any context and analysis. It cannot be useful itself unless we process to obtain knowledge.

In addition, knowledge is dynamics due to the huge social data generating by time. People even do not know that their activities on the internet fortuitously create knowledge. As a statistics from [2] in 2016, around 7000 tweets and 2,500,000 mails is sent; about 700 photos is uploads to Instagram in every second. Although one second is only a very short time in the real world, it means quite a lot on the internet. In [3], authors shows that sentiment of social media users towards blackberry, iphone, and android devices consecutive changes in a short period of time. Besides, authors mention that there are total 41 local events in Occupy Wall Street (OWS) at New York City during only few hours [4]. Due to aforementioned reason, we need an automated system to obtain our social knowledge rather than conduct manually.

Therefore, we build SocioScope as a framework with three core roles: $\left(i\right)$ collecting social data from different sources (e.g., social media, mass media, and sensor), $\left(ii\right)$ analyzing social data for understanding its relationships and connections to obtain social information, and $\left(iii\right)$ extracting hidden patterns to discover social knowledge [5]. From social knowledge, we can understand about societies around us. The scope of society is very dynamics and depends on the scope of social data that we get. Society can be a university, a company, or even a community.

In this section, we already gave the overview of our motivation. Section 2 provides the definition of Internet of Social Knowledge. Next, we will describe our system in Section 3. Further, the performance of SocioScope will be presented in Section 4. Finally, we conclude and discuss future work in Section 5.