The Yahoo! Shortcuts Module is great for evaluating words in a text string and categorizing them in Yahoo Pipes. It can create categories that include:
- PLACE
- PERSON
- IDENTIFIER (like phone numbers or email addresses)
- ORGANIZATION
- EVENT
- CONCEPT
After Yahoo! Shortcuts Module sorts the text strings into these categories, it includes sub-items that contain additional information about the category, such as: prediction probability, the start and end character locations where the item was found, and the section of the text that matched the category.
For example, if you type this string into Yahoo! Shortcuts:
”The Vancouver Canucks play hockey at General Motors Place.”
Vancouver Canucks” would be categorized as an ORGANIZATION, and “General Motors Place” would be categorized as a PLACE.
In this example, we’re going to type this sentence:
“Wayne Gretzky lit the Olympic torch in Vancouver, BC for the 2010 Olympics.” Into the Text Input Module and wire it into the Yahoo Shortcuts Module to “emit as a single item”.
“2010 Olympics” is categorized as an EVENT, “Olympic torch” is categorized as a CONCEPT, and “Wayne Gretzky” is categorized as a PERSON.
Each category is assigned a prediction probability score and a relevance score out of 100 (1 being the least accurate, and 100 being the most accurate).
The PLACE category provides geocode data that includes:
- State
- Country
- Latitude
- Longitude
- Zip code
…so you can further process it (such as plot the location on a map).
Here’s how are data looks when the results are emitted as a single item (drilled down):
And here’s how it looks when it’s emitted as a list (see how the keywords become the title?):
Yahoo Shortcuts! Is a great module for breaking up a text string to categorize each part and further manipulate it. For example, after names have been categorized, you could provide links to that person’s Wikipedia page or Twitter page to learn more about them. Or you could provide maps for locations that have been categorized as places.
Recent Comments