Data Conversion is defined as the modification of data structures to comply with new or different requirements for the data. Examples of this could include unpacking compressed fields or converting from EBCDIC to ASCII . Probably the two most common requirements for data conversion are database records and word processing systems. This is mostly because they make up the majority of computer applications and generally the way that data is stored.
Different operating systems have unique types of application software, and each application has a unique internal way of saving the data. There are some general standards such as RTF files for word processing text and CSV files for databases, however, these are few and far between, and they often only save the basic information rather than the complete data structure. Other major areas of divergence come from mainframe applications that use packed numbers and EBCDIC to keep data. Data conversion problems can mostly occur with any data interchange system, and although the Internet and networks can hide many format and media interchange problems, data incompatibility can still remain an issue.
Data conversions can also be distinguished by simple and complex. A simple data conversion can be defined as the transferring of files between different types of operating systems or backup routines without actually converting the data. Some examples of a simple data conversion are transferring data between different types of PC backups (Ex. Sytos Plus to NT Backup) or transferring data from DLT IV backup tapes to LTOs..
Complex data conversions generally involve converting the file structure. In most instances of a complex data conversion, you will write the converted data to a different tape, CD and/or operating system. Examples include conversion of DC600 System 36 files into Word files, convert EBCDIC database with binary and packed fields to an ASCII comma delimited or csv file for a PC, conversion of AS400 database in fixed length EBCDIC to an ASCII comma delimited or csv file for a PC, convert datasets and librairies from ICL and AS400 systems to ASA files for COM and COLD companies.
On many occasions you may not be aware that a data conversion has been done. Often in a Windows environment, a file is clicked on and an application opens and displays the information you require. Very often the original data is not in the native format of the application, and so a conversion is required. This may be fully automatic, or prompted for. It may because the data is from an earlier version of the program, (eg. Word 6 rather than Word 2000) or it has come from a completely different application. When the data conversion is clean and automatic, it will go unnoticed.
Moving data between very different systems can require a more actively involved data conversion procedure. This can be particularly true when moving information between mainframes and PCs or Unix systems. The two major differences are that mainframes tend to work in EBCDIC and have the concept of records, while PCs use ASCII and data is a continuous stream and not blocked in records. Data is encoded as EBCDIC with records being either fixed in length or with a binary length indicator. To complicate matters further, the numeric representations used within data from these systems is often packed decimal. In order for a data conversion to be successful you will need to take your COBOL record descriptions and use them to recreate your data in a usable ASCII form. You can then process your copy or save format data and convert it into delimited ASCII with printable numeric fields that are suitable to be loaded into most PC/UNIX applications.
Data conversion allows you to keep the same information but allow it to be processed by a different application, a different computer, or both. An example can be when information has to be moved to a different company or when computer systems are upgraded. It is also common when companies merge and want a new common computer system. Forensically there can be great advantages if the data can be converted into a format that can be processed with familiar tools, but in this case you must always be aware that any data conversion could possibly change the data and so must be approached with caution. However it can help with initial analysis of data.