Open Access
Review
Table 1.
Summary of behavioral identity authentication
| Object | Description | Characteristics | Issues |
|---|---|---|---|
| Keystroke [17, 77–83] | Keystroke refers to the action of a user inputting information through a keyboard, a keypad, or even a mouse. | (1) Without reliance on additional devices. (2) Extra layer of security for password. (3) Non-invasive continuous authentication. | Data quality easily affected by environmental and user conditions. |
| Touch gesture [3, 19, 84–87] | Touch gesture refers to the interaction between users and devices through gesture or touch. | (1) Covering multiple dimensional features to improve the security of verification. (2) Ensuring frictionless authentication. | (1) Difficulty in distinguishing and identifying multiple behavior patterns. (2) Cross-platform versatility limited by inconsistent behavioral data formats. |
| Motion [1, 4, 18, 88–91] | Motion refers to the various postures and movements made by users while using wearable or mobile devices. | (1) Utilization of broader behavioral features. (2) Non-invasive continuous authentication. (3) Less prone to variation caused by external factors. | (1) Multi-device dependency. (2) Privacy risks posed by sensor attacks. |
| Intrinsic signaling behavior [15, 92–94] | Intrinsic signaling behavior refers to the signals emitted by human organs during the interaction between users and devices. | High biometric uniqueness. | (1) High device dependency. (2) Data quality largely affected by emotions, diseases, etc. |
| User interaction behavior [12–14, 95–97] | User interaction behavior refers to the behavior of users interacting with applications. | (1) No requirement for sensor data collection and conversion. (2) A wider range of behaviors beyond keystroke or motion. | (1) Suffering from data privacy problem and breach risk. (2) The increasing probability of misjudgments of the model due to adversarial attacks. |
| Multi-factor [2, 98–100] | Multi-factor refers to the use of multiple categories of behavioral data or a combination of conventional methods with behavioral authentication to verify a users identity. | (1) Flexible combination of authentication methods. (2) The addition of cross-validation for an extra layer of security. (3) Reducing reliance on a single piece of sensitive data. | (1) The challenge of seamlessly integrating various behavioral features and authentication technologies. (2) Imbalanced data from various behavioral features. |
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.