Ansicht
Dokumentation

SMI_SENTIMENT_ASSIGNMENT - Report SMI_SENTIMENT_ASSIGNMENT

CPI1466 during Backup   Addresses (Business Address Services)  
This documentation is copyright by SAP AG.

Purpose

You use this report to assign a sentiment to an entity and copy search terms from the SOCIALDATA table into SMI_VOICE_CUST. The report ensures that a social post that is harvested more than once does not have duplicate entries in the SMI_VOICE_CUST table.

Integration

Prerequisites

• Index table $TA_INDEXVOICEOFCUST exists

• Text Analysis is up and running

Features

Sentiment Assignment:

Language filter options are available for sentiment assignment. There are three different processing scenarios:

• There is an entity but no sentiment has been associated with it

• There is an entity and it is possible to assign a sentiment

• A sentiment was found but it cannot be assigned to an entity

The report handles three scenarios:

Case 1:

The entity is there but no sentiment is found for the entity. The fields AVG_SENTIMENT and AVG_SENTIMENT_VAlUATION have the value zero.

Case 2:

An entity was found and it as possible to assign a sentiment to it. For example, for the post, "SAP HANA has a JavaScript Engine, SAP HANA performs text analysis. SAP HANA is great."

In this case, there are three references to the entity SAP HANA but two of those references have no sentiments associated with them. "SAP HANA is great" shows a strong positive sentiment. In this case, the average sentiment is 2 and the sentiment valuation is 5.

Case 3:

A sentiment is found but it cannot be assigned to an entity.

For example, no entity has been identified for the post "War is bad" but sentiment is associated with it.

Copying Search Terms:

You can copy search terms from the SOCIALDATA table into SMI_VOICE_CUST. An entry is created in the SMI_VOICE_CUST table even if the SOCIALPOSTSEARCHTERMTEXT field is empty in the SOCIALDATA table.

Selecting Entity Types to Process for Sentiment Assignment

You can select the entity types to be processed for sentiment assignment. For more information see Customizing node under Cross-Application Components->Process and Tools for Enterprise Applications->Social Intelligence->Text Analysis->Selecting Entity Types to Process for Sentiment Assignment.

Selection

Standard Variants

Output

Activities

The number of posts to be processed for Sentiment Assignment and Copying Search Terms is displayed on the screen.

To execute this report, proceed as follows:

Sentiment Assignment,,

1. Select the Process Sentiment Assignmentcheckbox to process a sentiment assignment.
2. Enter the planned number of packages.
3. Select the supported languages form the language filter option.

Copy Search Terms

1. Select the Copy Search Terms checkbox if you want to copy the search terms from the SOCIALDATAtable into SMI_VOICE_CUST.
2. Enter the planned number of packages.

General Settings

Note: You need to follow steps 6-9 for both Sentiment Assignment and Copying Search Terms.

1. Enter the date for post processing. If you do not enter a value in this field, the report is run for the period starting with the date it was last executed to the current date.
2. Enter the type of database connection. The database connection can be a local system (default) or secondary database. All WRITE operations are performed only in the local (Default) database. The READ operation is performed using the database connection mentioned in the Database Connection field. In case a secondary database connection is used, you must start the report in the client the posts have been harvested.
3. Select the Test Mode checkbox if you want to see the result and not store it in the back end.
4. Choose Executeor press (F8).
   

Example

PERFORM Short Reference   SUBST_MERGE_LIST - merge external lists to one complete list with #if... logic for R3up  
This documentation is copyright by SAP AG.

Length: 5185 Date: 20240601 Time: 061836     sap01-206 ( 72 ms )