The interpretation of dairy data using interactive visualization.

摘要

The agricultural domain is faced with the challenge of increasing amounts of collected data, multiple sources of external information and various summary reports, all tending to exist in a non-structured fashion. The formats and locations of these various data sources almost certainly differ, but profitable decision-making often depends on using and interpreting all of these inputs accurately. Based on the hypothesis that the interpretation of dairy-herd data can be aided using interactive visualization, the main goal of this research was to improve the understanding of the data-interpretation process, using such techniques, and by taking full advantage of them in the context of the Quebec dairy farm. The challenge was one of designing a system that would integrate interactive visualization techniques while also accounting for the diversity of information involved, the clientele of the dairy industry, and the technologies available. A methodological framework was developed to support the design of such an interactive visualization system, and a software prototype---HerdLine---subsequently developed. This prototype focused on the planning of a dairy-herd composition, based on the evolution of its performance over time via genetic and economic profiles. The framework helped the prototype development by better matching the characteristics of the targeted users with the data involved in the software. The result was a tool that provides an overall picture of the dairy-herd, as well as rapid, incremental, and reversible views of information contained within. An informal evaluation process was performed in order to assess the appropriateness of the prototype from a user's point of view. Four participants, representing diverse sections of the dairy industry, downloaded, installed and tested the prototype. They subsequently completed a survey regarding their impressions to such an approach, as well as their actual experience with the prototype. Both quantitative and qualitative responses were collected and documented. Comments obtained were generally positive and suggest that such a visual and interactive approach may be worthwhile, not only in the dairy industry, but also in other industries and sectors where decision-making is dependent on large, diverse sources of information.