4.6.6.3 Run Topic Modeling on Trained Text

After training the model, you can run topic modeling on the trained text.

To run topic modeling, use the sample data from the movies dataset. The dataset has no target column. When the output table generates, you can review the generated word groups and expressions for the trained text.

Before You Begin

Complete the following tasks:

• Prepare Data for Topic Modeling.

• Train a Model with Topic Modeling

Generating Topic Modeling for a Table

1. If not already done, load the model. You can use the session variable for the model that is valid for the duration of the connection. Alternatively, you can use the model handle previously set. For the option to set the user name, you can set it to NULL.

   The following example uses the session variable.

   mysql> CALL sys.ML_MODEL_LOAD(@model, NULL);

   The following example uses the model handle.

   mysql> CALL sys.ML_MODEL_LOAD('topic_modeling_use_case', NULL);

2. Run topic modeling on the dataset by using the ML_PREDICT_TABLE routine.

   mysql> CALL sys.ML_PREDICT_TABLE(table_name, model_handle, output_table_name), [options]);

   Replace table_name, model_handle, and output_table_name with your own values. Add options as needed.

   The following example runs ML_PREDICT_TABLE on the dataset previously created.

   mysql> CALL sys.ML_PREDICT_TABLE('topic_modeling_data.movies', @model, 'topic_modeling_data.topic_modeling_predictions', NULL);

   Where:

   • topic_modeling_data.movies is the fully qualified name of the input table that contains the data to run topic modeling for (database_name.table_name).

   • @model is the session variable for the model handle.

   • topic_modeling_data.topic_modeling_predictions is the fully qualified name of the output table with generated word groups and expressions (database_name.table_name).

   • NULL sets no options for the routine.

3. Query the ml_results column from the output table. Review the generated word groups and expressions for the movie descriptions next to tags.

   mysql> SELECT ml_results FROM topic_modeling_predictions;
   +-----------------------------------------------------------------------------------------------------------------------------------------+
   | ml_results                                                                                                                              |
   +-----------------------------------------------------------------------------------------------------------------------------------------+
   | {"predictions": {"tags": ["dangerous", "future", "join force", "journey", "warrior", "battle", "rebel", "force", "join", "group"]}}     |
   | {"predictions": {"tags": ["machine", "world", "agent", "group rebel", "human", "real", "real world", "power", "rebel", "fight"]}}       |
   | {"predictions": {"tags": ["family", "empire", "rival", "criminal", "powerful"]}}                                                        |
   | {"predictions": {"tags": ["scientist", "astronaut", "attempt", "humanity", "massive", "work", "earth", "ensure", "power", "survival"]}} |
   | {"predictions": {"tags": ["include", "mysterious", "strange", "world", "woman", "young woman", "young"]}}                               |
   | {"predictions": {"tags": ["empire", "force", "dark force", "evil", "group rebel", "join force", "wise", "battle", "dark", "help"]}}     |
   | {"predictions": {"tags": ["alien", "attempt", "humanity", "planet", "battle", "creature", "ensure", "scientist", "survival", "team"]}}  |
   | {"predictions": {"tags": ["future", "human", "creature", "form", "learn", "secret", "discover", "young"]}}                              |
   | {"predictions": {"tags": ["agent", "deal personal", "murder", "personal", "strange", "victim", "form", "help", "deal", "young"]}}       |
   | {"predictions": {"tags": ["wizard", "power", "dark", "learn", "school", "secret", "young", "destroy", "discover", "man"]}}              |
   | {"predictions": {"tags": ["alien", "dangerous", "massive", "mission", "planet", "work", "criminal", "earth", "team", "destroy"]}}       |
   | {"predictions": {"tags": ["murder", "student", "challenge", "help", "school", "face", "woman", "young woman", "young"]}}                |
   | {"predictions": {"tags": ["group", "fall", "friend", "place", "victim", "creature", "survival", "travel", "discover", "fight"]}}        |
   | {"predictions": {"tags": ["fall", "friend", "machine", "secure", "ensure", "scientist", "travel", "deal", "man", "young man"]}}         |
   | {"predictions": {"tags": ["place", "sister", "young", "fight", "woman", "young woman"]}}                                                |
   | {"predictions": {"tags": ["deal personal", "mysterious", "personal", "secure", "criminal", "powerful", "deal", "group"]}}               |
   | {"predictions": {"tags": ["wizard", "dark force", "evil", "include", "journey", "warrior", "wise", "dark", "powerful", "destroy"]}}     |
   | {"predictions": {"tags": ["family", "sister", "challenge", "deal", "woman", "young woman", "young"]}}                                   |
   | {"predictions": {"tags": ["astronaut", "mission", "earth", "face", "group"]}}                                                           |
   | {"predictions": {"tags": ["world", "real", "real world", "rival", "challenge", "discover", "face", "join", "man", "young man"]}}        |
   | {"predictions": {"tags": ["student", "form", "learn", "school", "secret", "force", "group"]}}                                           |
   +-----------------------------------------------------------------------------------------------------------------------------------------+
   21 rows in set (0.0472 sec)

To modify the number of word groups in the ml_results column, you can set the topk option. This option must be an integer greater or equal to one. The following example uses the same trained table and adds the option to limit the number of generated word groups to five.

mysql> CALL sys.ML_PREDICT_TABLE('topic_modeling_data.movies', @model, 'topic_modeling_data.topic_modeling_predictions_five', JSON_OBJECT('topk', 5));

Query the ml_results column to review the top five generated word groups.

mysql> SELECT ml_results FROM topic_modeling_predictions_five;
+----------------------------------------------------------------------------------------------+
| ml_results                                                                                   |
+----------------------------------------------------------------------------------------------+
| {"predictions": {"tags": ["dangerous", "future", "join force", "journey", "warrior"]}}       |
| {"predictions": {"tags": ["machine", "world", "agent", "group rebel", "human"]}}             |
| {"predictions": {"tags": ["family", "empire", "rival", "criminal", "powerful"]}}             |
| {"predictions": {"tags": ["scientist", "astronaut", "attempt", "humanity", "massive"]}}      |
| {"predictions": {"tags": ["include", "mysterious", "strange", "world", "woman"]}}            |
| {"predictions": {"tags": ["empire", "force", "dark force", "evil", "group rebel"]}}          |
| {"predictions": {"tags": ["alien", "attempt", "humanity", "planet", "battle"]}}              |
| {"predictions": {"tags": ["future", "human", "creature", "form", "learn"]}}                  |
| {"predictions": {"tags": ["agent", "deal personal", "murder", "personal", "strange"]}}       |
| {"predictions": {"tags": ["wizard", "power", "dark", "learn", "school"]}}                    |
| {"predictions": {"tags": ["alien", "dangerous", "massive", "mission", "planet"]}}            |
| {"predictions": {"tags": ["murder", "student", "challenge", "help", "school"]}}              |
| {"predictions": {"tags": ["group", "fall", "friend", "place", "victim"]}}                    |
| {"predictions": {"tags": ["fall", "friend", "machine", "secure", "ensure"]}}                 |
| {"predictions": {"tags": ["place", "sister", "young", "fight", "woman"]}}                    |
| {"predictions": {"tags": ["deal personal", "mysterious", "personal", "secure", "criminal"]}} |
| {"predictions": {"tags": ["wizard", "dark force", "evil", "include", "journey"]}}            |
| {"predictions": {"tags": ["family", "sister", "challenge", "deal", "woman"]}}                |
| {"predictions": {"tags": ["astronaut", "mission", "earth", "face", "group"]}}                |
| {"predictions": {"tags": ["world", "real", "real world", "rival", "challenge"]}}             |
| {"predictions": {"tags": ["student", "form", "learn", "school", "secret"]}}                  |
+----------------------------------------------------------------------------------------------+
21 rows in set (0.0475 sec)

To learn more about generating predictions for one or more rows of data, see Generate Predictions for a Row of Data.

What's Next

• Review other Machine Learning Use Cases.