0 / 0
Text Link Analyisis node (SPSS Modeler)

Text Link Analysis node

Sometimes, you might not need to create a category model to score. The Text Link Analysis (TLA) node adds a pattern-matching technology to text mining's concept extraction. TLA identifies relationships between the concepts in the text data based on known patterns. These relationships can describe how a customer feels about a product, which companies are doing business together, or even the relationships between genes or pharmaceutical agents.
Figure 1. Text Link Analysis node
Text Link Analysis node
  1. Add a Text Link Analysis node to your canvas and connect it to the Data Asset node that points to hotelSatisfaction.csv. Double-click the node to open its properties.
  2. Select id for ID field and Comments for Text field.
    Note: Only Text field is required.
    Figure 2. Text Link Analysis node FIELD properties
    Text Link Analysis node FIELD properties. It shows field settings like the ID field, Text field, Language field, Document Type, Textual Unity and Paragraph mode settings.
  3. For Copy resources from, select the Hotel Satisfaction (English) template.
  4. Under Expert, select Accommodate spelling for a minimum word character length of.
    Figure 3. Text Link Analysis node Expert properties
    Text Link Analysis node Expert properties. It shows check boxes for setting such as Accommodate spelling for a minimum root character limit, Extract uniterms, Extract nonlinguistic entities, Uppercase algorithm, Group partial and full person names together when possible, and Use derivation when grouping compound nouns.
    The resulting output is a table (or the result of an Export node).
    Figure 4. Raw TLA output
    Raw TLA output. It is a table with columns such as Concept1, Type1, Concept2, Type2, ID, and Matched text. Entries for concept columns are words such as room or parking. Entries for type columns are words such as Budget or Services. The rows show how a concept is related to a type or other concepts. Each row also shows how these words appear in the text.
    Figure 5. Counting sentiments on a TLA node
    Counting sentiments on a TLA node. It is a table with the columns ID, Comments, Pos_Count_Sum, and Neg_Count_Sum. Entries for the ID column are numbers for each row. Entries for the Comments column show short phrases extracted from the text. For example, one entry says Comfortable rooms, outstanding breakfast, and nice service. Entries for the Pos_Count_Sum, and Neg_Count_Sum columns show numbers counting the number of positive or negative sentiments for each short phrase. For example, for the previous phrase, it counted three positive sentiments.
Generative AI search and answer
These answers are generated by a large language model in watsonx.ai based on content from the product documentation. Learn more