Abstract:Temporal conceptual data modelling, as an extension to regular conceptual data modelling languages such as EER and UML class diagrams, has received intermittent attention across the decades. It is receiving renewed interest in the context of, among others, business process modelling that needs robust expressive data models to complement them. None of the proposed temporal conceptual data modelling languages have been tested on understandability and usability by modellers, however, nor is it clear which temporal constraints would be used by modellers or whether the ones included are the relevant temporal constraints. We therefore sought to investigate temporal representations in temporal conceptual data modelling languages, design a, to date, most expressive language, TREND, through small-scale qualitative experiments, and finalise the graphical notation and modelling and understanding in large scale experiments. This involved a series of 11 experiments with over a thousand participants in total, having created 246 temporal conceptual data models. Key outcomes are that choice of label for transition constraints had limited impact, as did extending explanations of the modelling language, but expressing what needs to be modelled in controlled natural language did improve model quality. The experiments also indicate that more training may be needed, in particular guidance for domain experts, to achieve adoption of temporal conceptual data modelling by the community.

What problem does this paper attempt to address?

This paper attempts to solve the following problems: 1. **Lack of user evaluation of Time Concept Data Modeling Language (TCDML)**: Existing TCDML has not been tested by users in terms of comprehensibility and usability, so it is not clear whether these languages meet the actual needs. 2. **Determine the best time representation and constraints**: Researchers hope to find an optimal representation of time elements and constraints, making the model easy to understand and use. 3. **Improve the quality of time - concept data models**: Through experimental verification, find out which factors can significantly improve the quality of time - concept data models. For example, using controlled natural language to describe the content to be modeled can improve the model quality. Specifically, the paper answers the following research questions through a series of experiments: - **RQ1**: Which graphical symbols do users prefer for representing time elements and constraints? - **RQ2**: How well do model builders understand the generated time - concept models? - **RQ3**: How capable are model builders in designing these models? To answer these questions, researchers conducted three groups of experiments: - The first group of experiments gradually determined the "best" time - information representation method through the qualitative reflective learning method. - The second group of experiments focused on model understanding and provided feedback on different symbolic variants of transformation constraints. - The third group of experiments mainly examined the modeling ability of model builders and carried out large - scale experiments by changing the level of detail in explanations and task requirements. Finally, these experiments helped researchers develop and improve a new time - concept data modeling language - **Trend**, and verified the effectiveness and ease of use of its graphical symbols in large - scale experiments.

The temporal conceptual data modelling language TREND

MERIT: Learning Multi-level Representations on Temporal Graphs.

Conceptual Temporal Modeling Applied to Databases

Managing temporal worlds for medical trend diagnosis

Remember This Event That Year? Assessing Temporal Information and Reasoning in Large Language Models

Language Models-enhanced Semantic Topology Representation Learning for Temporal Knowledge Graph Extrapolation

Temporal Modules

A Logical Temporal Relational Data Model

Towards Benchmarking and Improving the Temporal Reasoning Capability of Large Language Models

TRENDy: Temporal Regression of Effective Non-linear Dynamics

Unveiling Divergent Inductive Biases of LLMs on Temporal Data

The Development of Temporal Concepts: Linguistic Factors and Cognitive Processes

The current status of temporal network analysis for clinical science: Considerations as the paradigm shifts?

Towards Temporal Graph Databases

Analytical Models for Motifs in Temporal Networks: Discovering Trends and Anomalies

Preferences in Temporal Relational Databases

A Framework for Natural Language Interfaces to Temporal Databases

Systematic Theoretical Search for Dibaryons in a Relativistic Model

A survey for managing temporal data in RDF

Learning Temporal Causal Sequence Relationships from Real-Time Time-Series

An Overview Of Temporal Commonsense Reasoning and Acquisition