METICOS Deliverable D6.2 Social data analysis and extracted perceptions
Sarang Shaikh, Sule Yildirim Yayilgan, Mohamed Abomhara, Erjon Zoto · Zenodo (CERN European Organization for Nuclear Research) · 2022
Deliverable D6.2 provides state-of-the-art (SoTA) for the possible social media sources along with different types of social data as well as methods for gathering data from social media sources. Furthermore, this deliverable also provides SoTA review of the techniques available for analysing social media data. This deliverable links all of the SoTA discussions to the “perception extraction”, which is one of the key components involved in developing METICOS Social Sensing Toolkit for predicting technology acceptance of the automatic border control technologies. Beyond discussing the SoTA studies and perception extraction, this deliverable also discusses the perception extraction architecture pipeline designed for extracting perceptions for border control technologies from social media data as a part of T6.2. The preliminary implementations, experiments and results for perception extraction pipeline are also discussed in this deliverable (6.2). However, the more in detail discussions will be reported in the future deliverable (D6.3). The current deliverable report consists of eleven sections: Section 1 presents an introduction including the primary purposes, objectives and document outline. Section 2 provides a more general overview of the social media sources and social media data. Section 3 provides an in-depth review of different methods available for gathering social media data. Section 4 provides detailed overview as well as a SOTA review in AI techniques used for analysing social media data such as machine learning, deep learning and natural language processing. Section 5 and 6 discuss the perception extraction from the social media data and review the SoTA studies for perception extraction. Section 7 lists the major social media platforms where METICOS has an official group or page available for dissemination activities. Section 8 shows the proposed perception extraction architecture pipeline as a part of T6.2 to extract user perceptions for border control technologies along with architecture’s technical and implementation details as well. Section 9 discuss some of the preliminary results obtained with the proposed perception extraction architecture. Section 10 shows the mapping/relevance of the perception extraction results with METICOS Social Sensing Toolkit. Section 11 concludes the overall contribution of the deliverable. Due to the lack of bibliographical references about the implementation of perception extraction from social media data for border control technologies, D6.2 will be used as a guide for the following tasks of WP6, T6.2 (methods for gathering and analysing social media data) and T6.3 (METICOS Social Sensing Toolkit).