This paper focuses on some issues relating to data modelling, quality and management in a specific domain: forests. Many forest
domain specialists e.g., botanists, zoologists, economists and others collect vast volumes of data about the forest fauna
and flora, climate, soil, etc. The favourite tools for managing this data are spreadsheets and/or using popular DBMS packages
such as Access or FoxPro. The use of these tools introduces two major problems: loss of semantics and poor data structure.
These problems and associated issues are examined in this the paper.
To address these problems, we propose a method for database reverse engineering from spreadsheet tables to a conceptual model
and suggest a design of a prototype tool (TREAT). We also explain the motivation for and the methodology and approach that
we adopted. The interactive process used to identify the constituents of the spreadsheet tables and data semantics are explained.
Semi-automated analysis of the associations between the data items in terms of the domain knowledge, constraints and functional
dependencies between the data items are also described. The output from the tool may be selected as either an Entity-Relationship
or Object or Object-Relational model.
Keywords Data management - reverse engineering - data modelling