PaperlayerView exact revision on arXiv

Source-linked AI summary

The Internet of People: A human and data-centric paradigm for the Next Generation Internet

Marco Conti, Andrea Passarella

arXiv:2205.13970v1cs.NI

TL;DR

IoP still faces open design questions about primitive boundaries and resource trust. This opinion paper reviews data-management precursors that model human behavior and identifies research challenges; it concludes that those examples remain preliminary and need integration into a comprehensive framework.

  • Problem

    IoP design still faces open questions about which functions belong in basic versus enhanced primitives and how to represent, share, and trust node resources.

    Source: §2.1.2 · p. 9, §2.2.2 · p. 11

  • Method

    The paper outlines IoP’s characteristics and challenges and reviews existing data-management protocols that use quantitative models of individual and social behavior for filtering and dissemination.

    Source: §4 · pp. 18–19

  • Results

    Existing protocols embedding human-behavior models are preliminary examples of IoP and require significant refinement, improvement, and integration.

    Source: §4 · p. 19

  • Takeaways & Limitations

    A comprehensive IoP framework requires addressing open research challenges and integrating the preliminary systems that anticipate the paradigm.

    Source: §4 · p. 19

    It remains difficult to determine which functions belong in basic primitives and which in enhanced primitives.

    Source: §2.1.2 · p. 9

Abstract

from arXiv · show

The cyber-physical convergence, the fast expansion of the Internet at its edge, and tighter interactions between human users and their personal mobile devices push towards a data-centric Internet where the human user becomes more central than ever. We argue that this will profoundly impact primarily on the way data should be handled in the Next Generation Internet. It will require a radical change of the Internet data-management paradigm, from the current platform-centric to a human-centric model. In this paper we present a new paradigm for Internet data management that we name Internet of People (IoP) because it embeds human behavior models in its algorithms. To this end, IoP algorithms exploit quantitative models of the humans' individual and social behavior, from sociology, anthropology, psychology, economics, physics. IoP is not a replacement of the current Internet networking infrastructure, but it exploits legacy Internet services as (reliable) primitives to achieve end-to-end connectivity on a global-scale. In this opinion paper, we first discuss the key features of the IoP paradigm along with the underlying research issues and challenges. Then, we present emerging data-management paradigms that are anticipating IoP.

1. Introduction

Edge growth, data-centric use, and cyber-physical convergence make people central to the Next Generation Internet. The section proposes IoP as a human-centric shift in data management, with users’ devices acting on users’ behalf.

1. Introduction

  • The Internet’s edge is expanding as personal and IoT devices proliferate, while data use and cyber-physical interactions increasingly connect the Internet with the physical world.Users and devices are immersed in data that often has a strong local dimension.Source: §1 · pp. 1–2
  • Global platforms’ data-management role is associated with concerns about trust, privacy control, behavioral monitoring, relevance errors, and moving nearby data through centralized services.The paper argues these issues arise largely from platform-centric rather than human-centric data management.Source: §1 · pp. 2–3
  • IoP proposes placing people and their personal devices at the center of data management while using, rather than replacing, legacy Internet networking services.IoP functions would work above those services and dynamically select suitable communication options.Source: §1 · p. 3
  • Users’ devices become active proxies that autonomously build and configure data-management services through local or collaborative decisions, rather than leaving those tasks to centralized platforms.The devices are described as gateways through which users access the converging cyber-physical world.Source: §1 · p. 3
  • IoP algorithms on users’ devices would use quantitative models of individual and social behavior to reproduce how users make data-management choices.The paper distinguishes this direct behavioral correspondence from conventional bio-inspired approaches.Source: §1 · p. 3
  • Illustrative functions include relevance filtering with cognitive heuristics, social diffusion estimates using relationship models, and cognitive pruning before compact-DNN analysis.The paper identifies suitable abstractions, models, and mechanisms for IoP primitives and algorithms as key research challenges.Source: §1 · pp. 3–4, Figure 2 · p. 4

2. IoP Research Challenges

IoP research challenges span defining a human-centric data-management framework and primitives, modeling its graph and resources, and addressing privacy, analytics, and human behavior. IoP operates as an overlay on legacy Internet services, with graph relationships and data-management functions designed around people.

2. IoP Research Challenges

  • IoP overlays legacy Internet connectivity with a graph of devices, objects, and human relationships, and data-centric primitives for managing their interactions.The paper identifies graph and primitive definition, data collection and analytics, privacy, resource management, and human-behavior modeling as open research challenges.Source: §2 · p. 4, §2.1 · p. 4, §2.1.1 · p. 4

2.1.1 IoP graph

  • The IoP graph layers Internet connectivity, device-to-device links, human-device associations, and inherited social ties; movement and changing associations make it dynamic.Objects can inherit the social relationships of associated humans, while nodes may have only partial or imprecise graph knowledge.Source: §2.1.1 · pp. 5–7
  • Graph research must characterize both global and local properties, including robustness and information diffusion, and develop models usable for protocol configuration or graph modification.Complex-systems and control-theoretic approaches, including methods for partial or uncertain knowledge, are identified as relevant research directions.Source: §2.1.1 · p. 8

2.1.2 IoP primitives

  • IoP primitives are divided into basic functions and enhanced combinations, while both rely on legacy Internet primitives to move data between nodes.Data filtering, access, exchange, and privacy are candidate basic functions; dissemination can combine advertising, sending, and filtering.Source: §2.1.2 · pp. 8, 10
  • IoP primitives should use human behavior models for access, filtering, sharing, advertising, and trust, while embedding trustworthy data access and privacy-preserving policies.The boundary between basic and enhanced primitives remains an open design challenge.Source: §2.1.2 · p. 9

2.2 IoP data management and analytics

  • Data-management algorithms may reactively filter, proactively advertise or seek data, disseminate information, or analyze data, while operating over heterogeneous resources and partial graph knowledge.Open questions include which behavioral models to use, how to represent and orchestrate resources, how to preserve privacy, and how to design effective and efficient algorithms.Source: §2.2 · pp. 10–11

2.2.1 Human-centric models

  • IoP algorithm design draws on social science, cognitive psychology, complex-network analysis, and microeconomics to model relationships, trust, data interaction, and resource sharing.The paper calls for interdisciplinary development and validation, with Big Data Analytics and Machine Learning helping validate and tune models in IoP.Source: §2.2.1 · p. 11

2.2.2 Resources management and trust models

  • Resource-sharing models must account for dynamic, heterogeneous resources and social relationships, because self-interest alone does not correctly model cooperation patterns.Microeconomic models that incorporate social ties are proposed to describe access to shared resources and inform collaboration incentives.Source: §2.2.2 · pp. 11–12

3. First Steps towards IoP

The section surveys early data-management approaches that anticipate IoP, from human-centric collection to dissemination shaped by social relationships. These examples use personal devices and models of human behavior to manage information in cyber-physical settings.

3. First Steps towards IoP

  • People-centric networking uses nearby devices to exchange context-relevant data, while people-centric sensing recruits phones and other personal devices to collect physical-world information.Personal devices can also pool resources for opportunistic computing; the paper identifies these paradigms as precursors before focusing on data collection and dissemination.Source: §3 · pp. 12–13

3.1 Human-centric data collection

  • Cognitive heuristics address information overload by letting personal devices select relevant data using human-like, rapid judgments under partial knowledge.In, recognition-based ranking selects data for limited caches when devices meet, using local information rather than a full ranking that is typically infeasible.Source: §3.1 · pp. 14–15
  • Semantic networks model concepts and their links so that shared concepts can guide exchanges of concepts and associated data, while updating users’ interests dynamically.The approach in represents concepts rather than data items; starts exchanges from concepts shared by two devices.Source: §3.1 · p. 15

3.2 Human social relationships and data dissemination

  • Ego-network models use relationship structure and tie strength to inform data-dissemination protocols and assess how information spreads through online social networks.The review describes the ego-network model in and reports diffusion findings from.Source: §3.2 · pp. 15–17
  • Dunbar’s ego-network model groups alters into nested circles of differing tie strength, from a five-person support clique to an active network averaging 150 alters.The model also describes a 15-person sympathy group and a 50-person affinity group; Figure 6 depicts alters in concentric layers around the ego.Source: §3.2 · p. 16, Figure 6 · p. 16
  • Only 3% of nodes receive information when dissemination uses only very strong ties; adding even one below-threshold tie per ego can drastically increase spread.Trust-driven paths also have average shortest lengths more than twice those obtained when all social links are available.Source: §3.2 · p. 17

4. Discussion and Conclusions

The discussion frames IoP as a human- and data-centric Internet data-management paradigm, with personal devices acting as user proxies and human behavior embedded in system design. It identifies early systems as preliminary and calls for their integration and further development.

4. Discussion and Conclusions

  • IoP places personal devices at the center of data management as proxies for users, who retain control of their data and benefit from its collection and processing.The paper also envisions humans contributing cognitive resources and serving as network nodes when suited to particular algorithmic tasks.Source: §4 · pp. 17–18
  • IoP combines human-centered, multidisciplinary behavioral models with device-centric nodes, data-oriented functions, and self-organizing local networks.The paper calls for quantitative models grounded in fields including sociology, anthropology, psychology, economics, and physics.Source: §4 · p. 18
  • IoP is a data-management paradigm, not an application-level or policy framework, and it relies on the existing Internet for global connectivity when appropriate.Legacy Internet services remain enabling primitives rather than being replaced by IoP.Source: §4 · p. 18
  • IoP’s design challenge is embedding quantitative or algorithmic human-behavior models in protocols and device logic, rather than relying on qualitative descriptions.The paper distinguishes this approach from bio-inspired designs based on analogies with natural systems.Source: §4 · p. 19
  • Existing IoP precursors are preliminary and have mostly studied relevant elements in isolation, so the paper calls for refinement and integration in a comprehensive framework.Open challenges span the IoP graph and primitives, algorithms and protocols, resource management, trust, and privacy-preserving systems.Source: §4 · p. 19
Loading 2205.13970v1…