Analyses of temporal graphs provide valuable insights into temporal data through the use of two analytical approaches: temporal evolution and temporal information diffusion. The former shows how a network evolves over time; the latter explains how information spreads throughout a network over time. Systems have been mainly proposed to efficiently handle graph snapshots, which are suitable for temporal evolution but inappropriate for temporal information diffusion. For analyses of temporal information diffusion, temporal graph traversal platforms have recently been proposed; however, it is still infeasible to handle infinitely evolving temporal data, especially for monitoring applications. In this paper, we propose an incremental approach and its graph processing engine, Kairos, to enable prompt monitoring of temporal information diffusion. This approach makes it possible to immediately process diffusion results for sources of interest by traversing a part of the whole network, which avoids full traversals influenced by a small change in the network, thus making monitoring applications feasible. The recipes for implementing incremental versions of existing temporal graph traversal algorithms and metrics will make it easier for users to build their ad-hoc programs.

Analyses of temporal graphs provide valuable insights into temporal data through the use of two analytical approaches: temporal evolution and temporal information diffusion. The former shows how a network evolves over time; the latter explains how information spreads throughout a network over time. Systems have been mainly proposed to efficiently handle graph snapshots, which are suitable for temporal evolution but inappropriate for temporal information diffusion. For analyses of temporal information diffusion, temporal graph traversal platforms have recently been proposed; however, it is still infeasible to handle infinitely evolving temporal data, especially for monitoring applications. In this paper, we propose an incremental approach and its graph processing engine, Kairos, to enable prompt monitoring of temporal information diffusion. This approach makes it possible to immediately process diffusion results for sources of interest by traversing a part of the whole network, which avoids full traversals influenced by a small change in the network, thus making monitoring applications feasible. The recipes for implementing incremental versions of existing temporal graph traversal algorithms and metrics will make it easier for users to build their ad-hoc programs.

Temporal graph traversal is an approach for analyzing how information spreads throughout a network over time. A system has been recently proposed as an initial effort for efficient analyses against higher time complexity and infinitely evolving data unlike static graph. However, with the system, the response time for traversals from multiple sources is proportional to the number of sources; thus, application domains of the system can be limited. To resolve this problem, the state-of-the-art vertex-centric paradigm can be considered; however, we have found that the paradigm is not fitted into this computation. The paper proposes a novel time-centric computation approach for efficient all-pairs temporal graph traversals. One benefit of this approach is that users only need to focus on designing a repetitive task for graph elements that are valid at each sliding time, which simplifies the program logic and alleviates the burden of writing codes. Another benefit is that the approach is expected to enhance the performance by facilitating the reuse of intermediate results of multiple sources. The proposed approach is evaluated with a prototyped system, the recipes for existing algorithms, and the experiments with open temporal datasets. In addition, we also discuss how to handle ever-evolving real-world temporal networks.

Temporal graph traversal is an approach for analyzing how information spreads throughout a network over time. A system has been recently proposed as an initial effort for efficient analyses against higher time complexity and infinitely evolving data unlike static graph. However, with the system, the response time for traversals from multiple sources is proportional to the number of sources; thus, application domains of the system can be limited. To resolve this problem, the state-of-the-art vertex-centric paradigm can be considered; however, we have found that the paradigm is not fitted into this computation. The paper proposes a novel time-centric computation approach for efficient all-pairs temporal graph traversals. One benefit of this approach is that users only need to focus on designing a repetitive task for graph elements that are valid at each sliding time, which simplifies the program logic and alleviates the burden of writing codes. Another benefit is that the approach is expected to enhance the performance by facilitating the reuse of intermediate results of multiple sources. The proposed approach is evaluated with a prototyped system, the recipes for existing algorithms, and the experiments with open temporal datasets. In addition, we also discuss how to handle ever-evolving real-world temporal networks.

ChronoGraph is a novel system enabling temporal graph traversals. Compared to snapshot-oriented systems, this traversal-oriented system is suitable for analyzing information diffusion over time without violating a time constraint on temporal paths. The cornerstone of ChronoGraph aims at bridging the chasm between point-based semantics and period-based semantics and the gap between temporal graph traversals and static graph traversals. Therefore, our graph model and traversal language provide the temporal syntax for both semantics, and we present a method converting point-based semantics to period-based ones. Also, ChronoGraph exploits the temporal support and parallelism to handle the temporal degree, which explosively increases compared to static graphs. We demonstrate how three traversal recipes can be implemented on top of our system: temporal breadth-first search (tBFS), temporal depth-first search (tDFS), and temporal single source shortest path (tSSSP). According to our evaluation, our temporal support and parallelism enhance temporal graph traversals in terms of convenience and efficiency. Also, ChronoGraph outperforms existing property graph databases in terms of temporal graph traversals. We prototype ChronoGraph by extending Tinkerpop, a de facto standard for property graphs. Therefore, we expect that our system would be readily accessible to existing property graph users.

ChronoGraph is a novel system enabling temporal graph traversals. Compared to snapshot-oriented systems, this traversal-oriented system is suitable for analyzing information diffusion over time without violating a time constraint on temporal paths. The cornerstone of ChronoGraph aims at bridging the chasm between point-based semantics and period-based semantics and the gap between temporal graph traversals and static graph traversals. Therefore, our graph model and traversal language provide the temporal syntax for both semantics, and we present a method converting point-based semantics to period-based ones. Also, ChronoGraph exploits the temporal support and parallelism to handle the temporal degree, which explosively increases compared to static graphs. We demonstrate how three traversal recipes can be implemented on top of our system: temporal breadth-first search (tBFS), temporal depth-first search (tDFS), and temporal single source shortest path (tSSSP). According to our evaluation, our temporal support and parallelism enhance temporal graph traversals in terms of convenience and efficiency. Also, ChronoGraph outperforms existing property graph databases in terms of temporal graph traversals. We prototype ChronoGraph by extending Tinkerpop, a de facto standard for property graphs. Therefore, we expect that our system would be readily accessible to existing property graph users.