Electrical computers and digital processing systems: multicomput – Computer-to-computer data modifying
Reexamination Certificate
2001-10-10
2004-11-16
El-Hady, Nabil (Department: 2154)
Electrical computers and digital processing systems: multicomput
Computer-to-computer data modifying
C709S202000, C709S217000, C709S219000, C709S232000, C707S793000, C707S793000, C707S793000
Reexamination Certificate
active
06820135
ABSTRACT:
BACKGROUND OF THE INVENTION
Databases play an integral role in the information systems in most major organizations. Databases may take many forms, and play many roles such as a mailing list, an accounting spreadsheet, or statistical sales projections. After using several generations of technology, most organizations of any size have data stored in many different systems and formats. However, the increasing pace of competition is putting the onus on the organizations to build seamless bridges that combine the dizzying array of data sources quickly and cost-effectively into meaningful information.
In addition, increased opportunities brought by the World Wide Web (the “Web”) add to the pressure of providing access to that information in a useful and efficient manner. For example, organizations may need to transform raw data and stage it to separate, redundant web servers for quick access via multi-tier application architectures designed for thin clients. Alternatively, organizations may need to couple their systems via XML across the web with the systems of other organizations.
At the heart of nearly every Web-based business is the need to transform and integrate data. Because of the wide range of formats and applications within which business objects, transactions, catalog content and log files may be stored, data integration is perhaps the most painful and complex challenge facing business persons and application developers alike. The pain is most sharp in common scenarios that involve multiple trading partners with each partner having different internal production systems and different protocols for exchanging data.
Regardless of the particular need, organizations desiring to participate in any sort of e-commerce venture will likely have to deal with staging data from disparate sources. For example, an organization may need to combine information from multiple Internet systems along with external text feeds to build a customer relationship management system; or to integrate the organization's systems with those of customers and suppliers across the value chain. To be successful, an organization must transform data into useable formats for internal departments, partners, and customers.
To make data available and meaningful for different recipients, data transformation is often necessary. Data transformation generally refers to a sequence of operations that transforms a set of input data into a set of output data. Though the term data conversion has a slightly different technical connotation, it is often used synonymously with the term data transformation. Data transformation allows for the changing of the content, format, or data structure. Common content changes include adding, deleting, aggregating, concatenating, and otherwise modifying existing data. Common data formats include binary files, sequential files, embedded binary data, EBCDIC data from mainframes, common file types created by C, COBOL, FORTRAN, Basic, RPG, Pascal, and other languages, arrays, ISAMs and other record managers, PC-based databases, accounting applications, and Web-based data reachable through SQL/ODBC. Common data source structures may include spreadsheets, contact managers, mail list software, and statistical packages.
The process of converting data becomes increasingly complicated with each increase in the number of input data sources, the number of output data sources, the content of the data sources, the format of the data sources, and the complexity of data structures. For example, different data storage systems use data structures with different structures. For example, mainframe systems typically use a hierarchical data storage method, whereas client-server systems often use a relational database storage method.
Current data transformation techniques are generally expensive to implement, are not portable, and difficult to adapt to new or changing circumstances. For example, point-to-point links are generally hand-coded customized data transformation programs. Customized code is typically written in-house and is specific to a single application or DBMS environment. On the positive side, such solutions generally provide exactly what is needed and no more, and address requirements for which there may be no off-the-shelf products. In-house development, testing and debugging also narrows the focus, and tends to produce a workable, if non-versatile, solution. On the other hand, because these routines are usually specific to a particular source or target database, they are difficult to port to other environments. These routines may also be difficult to repeat because the routines are generally unique to each situation and because there is typically no infrastructure in place to manage the processes. Finally, building custom routines robs in-house DBAs of time better spent on other tasks. In addition, custom coded solutions require continued maintenance because they must be modified every time a new requirement is added to the system Further, custom code may take a relatively long time to implement with some legacy migration projects tying up critical IT staff for weeks, months and even years.
Consultants and customized tools are also used by organizations with increasing frequency today. Outside consultants typically have acquired extensive experience in building data models, designing movement and transformation methodologies and developing transformation tools. Such tools tend to be more portable, since they have been developed with multi-platform DBMS environments in mind. Because database consultants have had to become knowledgeable about business operations as well, these tools also tend to address business processes adequately. However, all application expertise leaves along with the consultant. In addition, because these routines are specific to single aspects of the business, they are difficult to recreate for other branches or divisions.
A common alternative to point-to-point links involves streaming data through a conduit into a universal structure, transforming the data in a central hub, then streaming the data through another conduit to the target format. Transforming the data may happen in real time but requires downloading the structure into memory to make it possible to apply a consistent set of visually defined transformation capabilities to an intermediate data stream regardless of the data's original format or transformed format. This method generally requires less programming code than point-to-point links. However, the architecture tends to limit transformation operations, and less efficient versions increase latency and complicate scalability. Further, loading the entire source structure into memory may become prohibitive with large and complex structures.
The use of an iterative method addresses this last concern. Iterative data transformation methods do not require the use of large amounts of memory that loading the entire structure requires, because source data is examined one record at a time. However, with current iterative methods, it is difficult to keep track of the relationship between records.
Another relatively new technology, XML, has lured application developers with the promise of an easier way to integrate data between applications and between organizations over the Internet. However, as organizations rush to adopt progressive e-business infrastructures such as XML to gain an edge over the competition, they are stumbling upon an unsettling reality. Since the W3C released the public specifications for XML in 1998, vertical industries and major corporations have already implemented hundreds of disparate XML “standards.” Thus, a big hurdle for trading partners, developers, and net market makers that need to interface with multiple production systems and organizations is the wide range of XML standards (as well as other data formats) that they will likely encounter.
Data transformation tools currently in use are generally expensive, time-consuming to implement, programming-intensive, and inflexible. An ideal data transformation tool requires a minimum amou
Bunton William G.
Dingman Paul C.
Landrum Thomas
Van Dyken Kathryn E.
Conley & Rose, P.C.
El-Hady Nabil
Huston Charles D.
Lettang Mollie E.
Pervasive Software, Inc.
LandOfFree
Modeless event-driven data transformation does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Modeless event-driven data transformation, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Modeless event-driven data transformation will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3282987