Before You Begin

This 15-minute tutorial shows you how to create a sentiment classifier preference and define its attributes with Oracle Text.

Background

A sentiment classifier is a type of document classifier that you use to extract sentiment metadata about a topic or document. To perform sentiment analysis by using a sentiment classifier, you must associate a sentiment classifier preference with the sentiment classifier.

What Do You Need?

• Oracle Database 18c
• Oracle SQL Developer 17.3.2

Create a Database Connection

1. Run sqldeveloper.exe in Windows or sqldeveloper.sh in Linux.
2. On the Connections tab, right-click Connections and select New Connection.
3. In the New / Select Database Connection window, enter or select the values for your database connection. This example uses the following values:
• Connection Name: sys_conn
• Username: sys
• Password: your SYS user password
• Role: SYSDBA
• Hostname: localhost
• Port: 1521
• SID: orcl
4. Click Test and verify the status of the connection. In this example, the status displays Success above the Help button.

   

5. Click Save, click Connect, and close the window.

When the connection is established, a SQL worksheet opens and you use it to run SQL commands.

Connect to an Oracle Text User

1. Create the user.

   CREATE USER myuser IDENTIFIED BY your_password;

2. Grant roles to the user.

GRANT RESOURCE, CONNECT, CTXAPP TO myuser;

The CTXAPP role enables you to create Oracle Text indexes and index preferences and to use PL/SQL packages.

3. Grant myuser an unlimited quota in the USERS tablespace.

   ALTER USER myuser QUOTA UNLIMITED ON USERS;

4. To create a connection to myuser, click the Connections tab, right-click Connections, and select New Connection.
5. In the New / Select Database Connection window, enter or select the values for your database connection. This example uses the following values:
• Connection Name: myuser_conn
• Username: myuser
• Password: your user password
• Role: default
• Hostname: localhost
• Port: 1521
• SID: orcl
6. Click Test and verify the status of the connection. In this example, the status displays Success above the Help button.

   

7. Click Save, click Connect, and close the window.

When the connection is established, a SQL worksheet opens and you use it to run SQL commands.

Create a Sentiment Classifier Preference

1. Create a sentiment classifier preference of SENTIMENT_CLASSIFIER type.

   EXEC CTX_DDL.CREATE_PREFERENCE('clsfier_camera','SENTIMENT_CLASSIFIER')

2. Define the attributes of the sentiment classifier preference.

EXEC CTX_DDL.SET_ATTRIBUTE('clsfier_camera','MAX_FEATURES','1000')
EXEC CTX_DDL.SET_ATTRIBUTE('clsfier_camera','NUM_ITERATIONS','600')

MAX_FEATURES enables you to specify the maximum number of distinct features (words, stems, and themes) to be extracted to build a sentiment classifier. NUM_ITERATIONS enables you to specify the number of iterations for which the sentiment classifier runs.