Managing the Quality of Information

Information processing is a business process that(from the chamber of Commerce) but a human
resembles a normal production process witheye will notice that both refer to the same
familiar demands for managing both the quantitycompany.
processed as well as the quality of the output.As there are many other fields in which company
For many business processes there is ainformation can differ (between the two sources),
continuous pressure to increase the output. Therethe (batch) process can calculate a value to each
is also constant demand for quality which acts ascompany date that represent the level of
a brake on this main process. In the informationmatching, where a perfect match is represented
processing area this problem is solved by usingby a 100%.
two different types of processes; batch andThe output of search engines work in the same
online. The quality indicator is the mechanism thatway. Search outcomes are sorted according to
will define how much the output is lowered inthis "match indicator level;" the lower the indicator
order to increase the quality (of the information).the lower the quality of the match.
An example of how this is done in practice youThis type of indicator can be used as a selection
could imagine The Yellow Pages. The booksmechanism between batch and online activities.
contain a variety of information about companiesFor instance by using a rule that all matches with
an each month a book is published with a selectiona level below 60% should be controlled by an
of companies in a certain region. This cycleagent.
continues until the last region of the country isIn this way the quality of the output is managed;
handled after which the book publishing processthe batch process to increase the level of output,
starts all over again with a series for the nextthe online part with human interference to check
year. Publishing these books is a process thatand increase the quality.
requires quite some organizing; most important isThis same technique could be (and perhaps is)
that the information is correct. Yet companiesused by managing article websites. When an
(and company information) do change a lot. Toarticle gets submitted to the site there are a
maintain this information the company informationseries of check required, which could also be done
in the data base needs to be checked withby a batch process. These check have also a
information from third parties (for example frommatching mechanism in them where (parts of)
the chamber of commerce).the article is checked against existing content on
Efficiency is important when organizing thesethe Web. This could give a resulting match level
activities and information systems can help toindicating the probability of the article authenticity.
organize this by separating batch from onlineFor yet other examples think about the IRS; all
activities. Batch obviously is a completelytax contributors are assigned a credibility indicator
automated process, online is where humanthat is calculated in a batch. The indicator is
interaction is required.derived in matching various other sources (banks
For example, in the batch process the companyand other financial institutions) and the way in
information in the database can be compared withwhich the tax form is filled in.
this third party data. The batch provides aWhat remains to be done in these environments
selection of companies where the match (of theis to define the quality level; how much do you
two data sources) is less than 100 percent. Thisdedicate to batch processing and how match to
means that human interaction is required toonline processing. This allocation question is what
(visually) check whether the third partydefines much of the quality of the overall output.
information is significantly different from the baseIt is about the question; "at what level (of the
data. "Microsoft Corp" (the base data) on oneindicator) are you going to check?" That's all up to
hand will show a difference from "Microsoft Inc."you.