AI-ANNE: (A) (N)eural (N)et for (E)xploration: Transferring Deep Learning Models onto Microcontrollers and Embedded Systems

This working paper explores the integration of neural networks onto resource-constrained embedded systems like a Raspberry Pi Pico / Raspberry Pi Pico 2. A TinyML aproach transfers neural networks directly on these microcontrollers, enabling real-time, low-latency, and energy-efficient inference while maintaining data privacy. Therefore, AI-ANNE: (A) (N)eural (N)et for (E)xploration will be presented, which facilitates the transfer of pre-trained models from high-performance platforms like TensorFlow and Keras onto microcontrollers, using a lightweight programming language like MicroPython. This approach demonstrates how neural network architectures, such as neurons, layers, density and activation functions can be implemented in MicroPython in order to deal with the computational limitations of embedded systems. Based on the Raspberry Pi Pico / Raspberry Pi Pico 2, two different neural networks on microcontrollers are presented for an example of data classification. As an further application example, such a microcontroller can be used for condition monitoring, where immediate corrective measures are triggered on the basis of sensor data. Overall, this working paper presents a very easy-to-implement way of using neural networks on energy-efficient devices such as microcontrollers. This makes AI-ANNE: (A) (N)eural (N)et for (E)xploration not only suited for practical use, but also as an educational tool with clear insights into how neural networks operate.

Longitudinal Sentiment Analyses for Radicalization Research: Intertemporal Dynamics on Social Media Platforms and their Implications

This discussion paper demonstrates how longitudinal sentiment analyses can depict intertemporal dynamics on social media platforms, what challenges are inherent and how further research could benefit from a longitudinal perspective. Furthermore and since tools for sentiment analyses shall simplify and accelerate the analytical process regarding qualitative data at acceptable inter-rater reliability, their applicability in the context of radicalization research will be examined regarding the Tweets collected on January 6th 2021, the day of the storming of the U.S. Capitol in Washington. Therefore, a total of 49,350 Tweets will be analyzed evenly distributed within three different sequences: before, during and after the U.S. Capitol in Washington was stormed. These sequences highlight the intertemporal dynamics within comments on social media platforms as well as the possible benefits of a longitudinal perspective when using conditional means and conditional variances. Limitations regarding the identification of supporters of such events and associated hate speech as well as common application errors will be demonstrated as well. As a result, only under certain conditions a longitudinal sentiment analysis can increase the accuracy of evidence based predictions in the context of radicalization research.

Sentiment Analysis with R: Natural Language Processing for Semi-Automated Assessments of Qualitative Data

Sentiment analysis is a sub-discipline in the field of natural language processing and computational linguistics and can be used for automated or semi-automated analyses of text documents. One of the aims of these analyses is to recognize an expressed attitude as positive or negative as it can be contained in comments on social media platforms or political documents and speeches as well as fictional and nonfictional texts. Regarding analyses of comments on social media platforms, this is an extension of the previous tutorial on semi-automated screenings of social media network data. A longitudinal perspective regarding social media comments as well as cross-sectional perspectives regarding fictional and nonfictional texts, e.g. entire books and libraries, can lead to extensive text documents. Their analyses can be simplified and accelerated by using sentiment analysis with acceptable inter-rater reliability. Therefore, this tutorial introduces the basic functions for performing a sentiment analysis with R and explains how text documents can be analysed step by step - regardless of their underlying formatting. All prerequisites and steps are described in detail and associated codes are available on GitHub. A comparison of two political speeches illustrates a possible use case.

Analysing Social Media Network Data with R: Semi-Automated Screening of Users, Comments and Communication Patterns

Communication on social media platforms is not only culturally and politically relevant, it is also increasingly widespread across societies. Users not only communicate via social media platforms, but also search specifically for information, disseminate it or post information themselves. However, fake news, hate speech and even radicalizing elements are part of this modern form of communication: Sometimes with far-reaching effects on individuals and societies. A basic understanding of these mechanisms and communication patterns could help to counteract negative forms of communication, e.g. bullying among children or extreme political points of view. To this end, a method will be presented in order to break down the underlying communication patterns, to trace individual users and to inspect their comments and range on social media platforms; Or to contrast them later on via qualitative research. This approeach can identify particularly active users with an accuracy of 100 percent, if the framing social networks as well as the topics are taken into account. However, methodological as well as counteracting approaches must be even more dynamic and flexible to ensure sensitivity and specifity regarding users who spread hate speech, fake news and radicalizing elements.