[Paper Review] Transaction Logic with (Complex) Events

Ana Sofía Gomes, José Júlio Alferes|arXiv (Cornell University)|May 15, 2014

Logic, Reasoning, and Knowledge17 references3 citations

TL;DR

This paper proposes Transaction Logic with Events ($\mathcal{TR}^{ev}$), a non-monotonic extension of Transaction Logic ($\mathcal{TR}$) that integrates complex event detection with transaction execution in general Knowledge Bases. It shows that $\mathcal{TR}$ alone cannot guarantee response to all detected complex events during transaction execution, and solves this by extending $\mathcal{TR}$ with event-aware syntax, model theory, and executional semantics that ensure every complex event detected in a transaction triggers a corresponding response within the same transaction.

ABSTRACT

This work deals with the problem of combining reactive features, such as the ability to respond to events and define complex events, with the execution of transactions over general Knowledge Bases (KBs). With this as goal, we build on Transaction Logic (TR), a logic precisely designed to model and execute transactions in KBs defined by arbitrary logic theories. In it, transactions are written in a logic-programming style, by combining primitive update operations over a general KB, with the usual logic programming connectives and some additional connectives e.g. to express sequence of actions. While TR is a natural choice to deal with transactions, it remains the question whether TR can be used to express complex events, but also to deal simultaneously with the detection of complex events and the execution of transactions. In this paper we show that the former is possible while the latter is not. For that, we start by illustrating how TR can express complex events, and in particular, how SNOOP event expressions can be translated in the logic. Afterwards, we show why TR fails to deal with the two issues together, and to solve the intended problem propose Transaction Logic with Events, its syntax, model theory and executional semantics. The achieved solution is a non-monotonic extension of TR, which guarantees that every complex event detected in a transaction is necessarily responded.

Motivation & Objective

• To address the gap in existing reactive systems that cannot simultaneously detect complex events and execute transactions with full transactional guarantees.
• To show that standard Transaction Logic ($\mathcal{TR}$) fails to ensure response to all complex events detected during transaction execution.
• To design a new logic, $\mathcal{TR}^{ev}$, that combines complex event reasoning with transactional execution in a declarative, non-monotonic framework.
• To provide a formal model theory and executional semantics for $\mathcal{TR}^{ev}$ that ensures every detected complex event triggers a corresponding action within the same transaction.
• To demonstrate that $\mathcal{TR}^{ev}$ can express most SNOOP event operators and supports complex event patterns used in CEP systems.

Proposed method

• Extend $\mathcal{TR}$ with event-specific syntax and semantics to support complex event detection within transactional contexts.
• Define a model theory for $\mathcal{TR}^{ev}$ based on minimal models that capture event occurrences and transaction execution paths.
• Introduce executional entailment in $\mathcal{TR}^{ev}$ that evaluates formulas relative to minimal models of the program and paths of state transitions.
• Integrate event detection with transaction execution by ensuring that every complex event detected during a transaction is responded to within that transaction.
• Leverage the complex event detection algorithm from ETALIS and the execution model of $\mathcal{TR}$ to define a sound execution procedure for $\mathcal{TR}^{ev}$.
• Use oracles to separate state transitions and event occurrences from the logic, enabling clear distinction between actions and event occurrences in the execution path.

Experimental results

Research questions

• RQ1 Can Transaction Logic ($\mathcal{TR}$) natively express complex events, including those defined by SNOOP event operators?
• RQ2 Why does $\mathcal{TR}$ fail to guarantee that all complex events detected during a transaction are responded to within that transaction?
• RQ3 How can a logic be extended to simultaneously support complex event detection and transactional execution with full consistency guarantees?
• RQ4 What is the formal model theory and executional semantics for a logic that ensures every detected complex event triggers a corresponding transactional response?
• RQ5 Can the resulting logic, $\mathcal{TR}^{ev}$, be shown to be a non-monotonic extension of $\mathcal{TR}$ that preserves correctness and supports expressive event patterns?

Key findings

• Standard $\mathcal{TR}$ cannot guarantee that all complex events detected during a transaction are responded to within the same transaction execution.
• The proposed $\mathcal{TR}^{ev}$ logic successfully extends $\mathcal{TR}$ with a declarative, non-monotonic framework that ensures every complex event detected in a transaction triggers a corresponding response.
• $\mathcal{TR}^{ev}$ supports the expression of most SNOOP event operators and complex event patterns used in CEP systems.
• The executional entailment in $\mathcal{TR}^{ev}$ is defined via minimal models of the program and paths of state transitions, ensuring sound and complete response to detected events.
• For event-free, serial-Horn $\mathcal{TR}$ theories, executional entailment in $\mathcal{TR}^{ev}$ coincides with that in $\mathcal{TR}$, preserving backward compatibility.
• The logic provides a sound foundation for reactive systems that require both complex event processing and transactional integrity, such as airline rescheduling or real-time service adaptation.