Data analytics can be seen as a powerful tool for the fifth-generation (5G) communication system to enable the transformation of the envisioned challenging 5G features into a reality. In the current 5G architecture, some first features toward this direction have been adopted by introducing new functions in core and management domains that can either run analytics on collected communication-related data or can enhance the already supported network functions with statistics collection and prediction capabilities. However, possible further enhancements on 5G architecture may be required, which strongly depend on the requirements as set by vertical customers and the network capabilities as offered by the operator. In addition, the architecture needs to be flexible in order to deal with network changes and service adaptations as requested by verticals. This paper explicitly describes the requirements for deploying data analytics in a 5G system and subsequently presents the current status of standardization activities. The main contribution of this paper is the investigation and design of an integrated data analytics framework as a key enabling technology for the service-based architectures (SBAs). This framework introduces new functional entities for application-level, data network, and access-related analytics to be integrated into the already existing analytics functionalities and examines their interactions in a service-oriented manner. Finally, to demonstrate predictive radio resource management, we showcase a particular implementation for application and radio access network analytics, based on a novel database for collecting and analyzing radio measurements. INDEX TERMS 5G, architecture, data analytics, network slicing. I. INTRODUCTION The fifth generation (5G) mobile communications system is characterized by a wide-range of services grouped under three generic service types, namely, enhanced mobile broadband (eMBB), massive machine-type communications (mMTC), and ultra-reliable and low-latency communications (URLLC). The network slicing concept is introduced in 5G to address the various requirements from multiple vertical industries assuming a shared physical network infrastructure. A network slice can be customized according to the needs of vertical industries and services to be supported. Network slicing is a key pillar in 5G networks. The associate editor coordinating the review of this manuscript and approving it for publication was Kostas Psannis. The end-to-end (E2E) nature imposes domain-specific requirements that will span over multiple technical domains, i.e., radio access network (RAN), transport network, and core network (CN). In addition, 5G shall be supported by a management and orchestration (M&O) layer in order to meet defined service-level agreements (SLAs) for network slices of different nature. In 3rd generation partnership project (3GPP), four standard slice/service types (SSTs) have been introduced, namely, eMBB, mMTC, URLLC, and vehicleto-everything (V2X) SSTs, which aim to provide differen...