I need your help regarding one Customer requirement in
EDQ. Below are the details about Customer requirement.
Customer wants Web Services in EDQ and he want to connect
this Web Service to a Database Data (data uploading very frequently into this
So, Customer wants refreshed data in EDQ all the time.
There is not much to go on here, but assuming the need is for attached matching (duplicate prevention) on dynamically changing data, you should probably read the Real-Time Matching Guide in the online help from EDQ. Basic content of which is as follows.
The Business Services guide of EDQ-CDS may also be useful.
EDQ provides two different types of real time matching:
This topic provides a general guide to how real time matching works in EDQ.
Note: If you wish to use EDQ for real-time duplicate prevention and/or data cleansing with Oracle Siebel UCM or CRM, Oracle provides a standard connector for this purpose.
Real time duplicate prevention
EDQ's real-time duplicate prevention capability assumes that the data being matched is dynamic, and therefore changes regularly (for example, customer data in a widely-used CRM system). For this reason, EDQ does not copy the working data. Instead, the data is indexed on the source system using key values generated using an EDQ cluster generation process.
Real-time duplicate prevention occurs in two stages - Clustering, and Matching.
In the Clustering stage, EDQ receives a new record on its real-time interface and assigns it cluster keys using a process that has been enabled for real-time execution. It returns the cluster keys for the record using a Writer. The source system then receives this message, selects candidate records using a pre-keyed table, and feeds back to EDQ a set of all the records that share one of the cluster keys with the driving record.
Normally, it is advisable to generate many key values per record, as is also the case for batch matching.
In the Matching stage, EDQ receives the message containing the records that share a cluster key with the input record. It then uses a match processor, with all records presented as working records on a single input, to select the definite and possible matches to the new record from these candidate match records, and assign them a score that can be used to sort the matches. The matching results are then returned on a real-time interface, and this response is handled externally to determine how to update the system. Possible responses include declining the new (duplicate) record, merging the new record with its duplicate record(s), or adding the new record, but including a link to the duplicate record or records.
Real-time reference matching
EDQ's real-time reference matching implementation matches new records against one or many reference sets. The data in the reference sets is assumed to be non-dynamic. That is, they are updated on a regular basis, but not constantly accessed and updated by multiple users (for example watch lists, rather than CRM data). Reference matching is a single stage process. Incoming records from a working real-time source are matched against snapshots of reference records from one or more staged sources. A writer on the process then returns the output of the match processor back to the calling system. Note that the output returned may be any of the forms of output from the match processor. If you want only to link the new records to the reference sets, you can write back the Relationships output. If you want to enhance the new records by merging in data from the matching reference records, you can use the merge rules in a match processor, and write back the Merged (or Enhanced) data output.
Note: The Reference Data in a real-time matching process can be cached and interrogated in memory on the EDQ server if required. See Cache reference records for real-time processes for further information.
Preparing a process for real-time reference matching
For a real-time reference matching process to perform correctly, it must first be run in Prepare mode. This allows it to compare the inbound records quickly against the reference sets and issue a response. Running a process in Prepare mode will ensure that all the cluster keys for the reference data sets have been produced.
Re-preparing reference data
Real-time matching uses a cached and prepared copy of the reference data. This means that when updates are made to reference data, the snapshot must be re-run and the data re-prepared before the updates can be propagated to the matching process.
Re-preparing reference data involves:
Triggers, which can start and stop other jobs, can be used to create a job which includes all the above phases. A Stop, re-prepare and restart job would appear as follows:
This job consists of three phases. The first stops the real-time match process and re-runs the reference data snapshot. The second runs the match process in Prepare mode. The third re-starts the real-time match process in Interval mode.
Triggers are configured to run either at the beginning of a phase or at the end, so it would be possible to include the restart trigger at the end of the second phase, just as the stop trigger is placed at the beginning of the first phase. However, placing the restart trigger in its own phase means you can configure the restart phase to run only if the re-prepare phase is successful; see "Using Job Triggers" for further information.
Enabling real-time matching
In order to enable real time matching, you must configure a process containing a match processor, and use a single working input (to the match processor) only.
Real time processes may be run in Interval mode. In Interval mode, the processes run on a continuous basis, and write their results to the results database at regular intervals. See Execution Options.
A real-time process can be run in Normal mode, but in this case:
Real time consumer and provider interfaces must also be configured to communicate with an external system, using either JMS or Web Services.
Note: For real-time reference matching via a Web Service, you can use EDQ's Web Service capability to generate the real-time consumer and provider interfaces automatically. If you are using the EDQ Customer Data Services Pack, pre-configured Web Services are provided.