Quantum Computing
For decades, computer scientists have been aware of the limitations of the standard von Neumann architecture that is the foundation for most computers in current use. The sheer scope and complexity of the problems being tackled by computational technology in conjunction with the emergence of Big Data has necessitated the search for alternative architectures. Parallel computing is one such alternative and has been applied successfully to a myriad of problems in all areas of engineering and in the physical, biological, medical, mathematical, and social sciences. However, partially due to concerns over computer security and the possible breach of cryptographic mechanisms, more exotic computational paradigms have been proposed. The proposal that has garnered the most attention, research, and currently exists at a high level of development is quantum computing. Quantum computing, based on the theoretical principles of quantum mechanics, breaks from the binary code, and instead represents data and states in qubits. Like a binary value, a qubit has two states. However, unlike the former, qubits can be in the two states simultaneously. (Note: This seeming contradiction is possible through the principle of superposition, a fundamental property of quantum mechanics. However, a detailed discussion of this and other issues related to quantum mechanics is beyond the scope of the current discussion.) This property of coherently being in two states simultaneously is what endows quantum computing with its power. Without delving into the (quite complex) physical and mathematical minutiae of quantum physics, it suffices to point out that quantum computing opens a vast array of formerly unsolvable problems (on von Neumann machines) to computational analysis. Machine learning, including problems in classification, clustering, or dimension reduction that employ computationally intensive optimization techniques, can particularly benefit from the quantum computing paradigm (Barzen, 2021). Not only can new problems be solved, but computational tasks (theoretically) can be performed much faster, with greater precision, greater energy efficiency, and with less expense than on traditional architectures (Barzen, 2021). It is important to note that quantum computing (as of 2021) is in the development stage and is not accessible for general purpose applications. Although some quantum computing devices with limited functionality exist and are commercially available, a large amount of work in the field is still at the theoretical level.
A use case from media science provides an example of the potential benefits of quantum computing for analyzing humanities data. A data analysis pipeline consisting of various data preprocessing techniques, feature extraction and feature engineering (determining and generating features suitable for analysis approaches), clustering (placing features into categories, or groups), and classification (determining to which category data belong, based on their features) are proposed for quantum computer implementation, as well as realization on standard von Neumann architectures. The goal of the research initiative, Project MUSE, is to identify a pattern language for costumes that appear in films (Barzen, 2021). Relevant information about costumes is acquired and analyzed to obtain important information into costume patterns. These costume patterns represent such aspects as colour, material, or manner in which the costume is worn, and convey a particular stereotype, character trait, or other relevant characteristics. It is proposed that quantum computing, while marking a significant advance in computational algorithms like machine learning, can more importantly open completely new possibilities for humanistic inquiry (Barzen, 2021).
