Talk:Data integration

the concepts of integrated data processing system

Removing commercial links

I removed the entire section of external links under the "Commercial" heading. This section was attracting spam and was in violation of WP:EL and WP:NOT - WP is not a link directory of commercial products. It did not improve the content of the article and it would be impossible to moderate if left in place. If you feel like any of these links were removed in error, please discuss their inclusion here before relinking. Thanks. Nposs 21:16, 19 January 2007 (UTC)

• ref. Nr. 8 is a commercial brochure and not about enhanced modeling methodology

Should not be merged

I am concerned about the possibility of merging this with the general data integration entry - Edge Data Integration, while related and similar in some ways is very different from what most folks consider data integration - different purpose, different tools, different patterns, different data. 71.56.69.30 (talk) 10:38, 22 June 2008 (UTC)

Removal of external link

I've removed an external link to Costs of Data Integration because I have serious concerns that it represents a conflict of interest. The external article seems to suggest cost savings available from using products by Pervasive Software, however one of the authors of the article might very well be the same as a user here on wikipedia who has stated that they worked for Pervasive for many years ( I'm not sure if I'm allowed to state the author and username involved because of policy on outing, will provide the further details to admins if wanted) and then the article is added to this page by Shaw76 (talk • contribs) who has openly stated here that they are an offically authorized spokesperson for Pervasive. TurningWork (talk) 19:16, 21 April 2009 (UTC)

Insufficient Inline References and Possibly Original Research

The "History" section provides a chronology (as of 2009...as of 2011) without providing any citations - the most recent citation in the Bibliography (other than Lane's, which is strictly a news item) is 2002.

An additional concern is that the article appears to set up straw men - thus, the complaint that a centralized data warehouse can go out of date is of less importance in the situation of a single organization, where single-vendor RDBMS technologies that allow near-real-time incremental data updates to materialized views can be deployed. While federated approaches using data mediation and a virtual schema are viable, they are preferentially employed only in those circumstances where the individual local data sources are developed independently of the team that manages the integration task (as in research consortia, where the maintainers of individual sources collaborate - often loosely - but do not give up autonomy). In such conditions, physical data integration is often not politically feasible.

The classic paper of Won Kim et al on the challenges of integrating heterogeneous schemas (Won Kim, Injun Choi, Sunit Gala, and Mark Scheevel. On resolving schematic heterogeneity in multidatabase systems. Distributed and Parallel Databases, 1(3):251–277, July 1993) should be cited .

Prakash Nadkarni (talk) 02:11, 4 May 2012 (UTC) May 3 2012