The field of Information Theory is founded on Claude Shannon's seminal ideas relating to entropy. Nevertheless, his well-known avoidance of meaning (Shannon, 1948) still persists to this day, so that Information Theory remains poorly connected to many fields with clear informational content and a dependence on semantics. Herein we propose an extension to Quantum Information Theory which, subject to constraints, applies quantum entanglement and information entropy as linguistic tools that model semantics through measures of both difference and equivalence. This extension integrates Denotational Semantics with Information Theory via a model based on distributional representation and partial data triples known as Corolla.
The field of Information Theory is founded on Claude Shannon's seminal ideas relating to entropy. Nevertheless, his well-known avoidance of meaning (Shannon, 1948) still persists to this day, so that Information Theory remains poorly connected to many fields with clear informational content and a dependence on semantics. Herein we propose an extension to Quantum Information Theory which, subject to constraints, applies quantum entanglement and information entropy as linguistic tools that model semantics through measures of both difference and equivalence. This extension integrates Denotational Semantics with Information Theory via a model based on distributional representation and partial data triples known as Corolla.
Information-theoretic measures such as relative entropy and correlation are extremely useful when modeling or analyzing the interaction of probabilistic systems. We survey the quantum generalization of 5 such measures and point out some of their commonalities and interpretations. In particular we find the application of information theory to distributional semantics useful. By modeling the distributional meaning of words as density operators rather than vectors, more of their semantic structure may be exploited. Furthermore, properties of and interactions between words such as ambiguity, similarity and entailment can be simulated more richly and intuitively when using methods from quantum information theory.
This paper introduces a novel approach to information theory, leveraging quantum entanglement entropy as a fundamental measure of information. We propose that entanglement entropy can effectively quantify the efficiency of information transmission and provides a theoretical foundation that surpasses the limitations of classical information theory. The core of this work lies in the integration of entanglement entropy with Shannon entropy, establishing a new framework for analyzing information transfer. We explore scenarios where classical information theory struggles to adequately describe the limits of communication due to quantum mechanical effects, specifically entanglement. The paper aims to provide a theoretical underpinning for quantum communication protocols and potentially inform the development of quantum computation. We define key terms and formulate core concepts, focusing on the relationship between entanglement entropy, information content, and the constraints imposed by quantum mechanics. The central argument is that by considering the inherent interconnectedness of entangled systems, we can gain a deeper understanding of information processing at the most fundamental level. ---
Information theory, introduced by Claude Shannon in 1948, provides a mathematical model, predominantly probabilistic, for addressing the challenges associated with measuring the amount of information in a message and efficiently and reliably storing the message (compression and protection). This theory has always been intricately connected to communication theory. Information theory not only relies on communication theory but also draws from fields such as statistics, statistical physics, and inference theory. In this chapter, a concise overview is presented, covering the mathematical foundations of classical and quantum information, their manipulation and associated qualifications. The fundamental concepts of classical information theory, including Shannon entropy, joint entropy, conditional entropy, and mutual information, are introduced. Additionally, basic concepts related to quantum information and quantum correlation measurements are presented.
The idea that information can be localized is pervasive in physics. In thermodynamics, entropy flows from one reservoir to another, and in quantum gravity, the entropy content of some spatial regions is bounded. In quantum information theory, information is transmitted from one place to another, and entanglement can provide an advantage when there are constraints on such transmission.
Shannon entropy quantifies localizable information in the sense that marginalization defines an outer measure on the set of parts of a classical multipartite system. In contrast, von Neumann entropy does not quantify localizable information in this sense. However, a variant of von Neumann entropy, which originates in noncommutative geometry, does.
We prove that this adjusted von Neumann entropy is the minimum quantum entropy that quantifies localizable information. We also prove that this is the unique quantum entropy that characterizes Bell states in terms of redundant information, directly generalizing the classical case. We work with finite-dimensional C∗-algebras throughout, modeling finite quantum systems that may have superselection sectors.
Classical information theory, built on Shannon entropy, and quantum information theory, built on von Neumann entropy, quantify uncertainty and compression properties of probability distributions and density operators. Both are powerful, but both treat information as a property of states of individual systems or ensembles. In this manuscript we argue that entanglement—the pattern of non-classical correlations between systems—is more fundamental than either Shannon or von Neumann information. We motivate this claim with a condensate thought experiment in which many bosons occupy the same single-particle state: Shannon and von Neumann entropies assign them identical information content, yet they can be distinguished by their distinct entanglement histories with external systems. Formally, states with identical marginal density operators but different purifications represent different global information, accessible only through their entanglement structure. Entropic quantities are then seen as coarse-grained summaries of this underlying relational information. We connect this perspective to an entanglement-geometric view of physics, including Entanglement Manifold Cosmology (EMC), in which spacetime and physical “truth” emerge from the geometry of correlations. The goal is an information-theoretic narrative that is rigorous enough for physics and information theory while readable to a non-specialist who is comfortable with basic linear algebra and quantum mechanics.
Classical information theory relies on the probabilities that signals appear in a source. As such, information depends on Kolmogorov’s probability theory. However, it is well known that some quantum sources are context-dependent, and therefore may not conform to classical probability theory. For such sources, information can be defined through von Neumann’s entropy. However, there are possible contextual sources that are neither quantum nor classical. In this chapter, we explore a simple example of one such source, and show that, similarly to the quantum case, the presence of contextuality increases the amount of information. We also show that, by extending Shannon’s entropy to accommodate negative quasi-probability distributions, we are able to provide a measure of informational content that is neither described by Shannon nor von Neumann’s entropies.