UIMA Ruta中能否通过WORDTABLE实现文本部分匹配？

Absolutely! You can pull off this suffix-based matching and feature assignment using UIMA Ruta's WORDTABLE functionality. Here's a step-by-step breakdown tailored to your specific case:

1. Prepare Your CSV Mapping File #

First, confirm your CSV file (let’s name it suffix_mappings.csv) uses the correct semicolon separators as you provided:

book;b.
ship;sh.

Each row links a target suffix to its corresponding feature value.

2. Configure WORDTABLE for Suffix Matching #

By default, WORDTABLE uses prefix matching, but Ruta lets you switch to suffix mode with simple parameters. Add this declaration to your Ruta script:

// Declare the WORDTABLE with suffix matching enabled
WORDTABLE suffixMap = "suffix_mappings.csv";
WORDTABLE(suffixMap, false, ";", 1, 2, SUFFIX);

// Enable case-insensitive matching by changing the second parameter to true:
// WORDTABLE(suffixMap, true, ";", 1, 2, SUFFIX);

Let’s break down those parameters:

• false: Disables case-insensitive matching (flip to true if you need to match "Book" or "SHIP" variants)
• ;: The delimiter used in your CSV file
• 1: Column index (starting at 1) for suffix keys (book, ship)
• 2: Column index for feature values (b., sh.)
• SUFFIX: Tells Ruta to match the end of each word against the table’s keys

3. Apply the Matching & Feature Assignment Rule #

First, make sure you’ve defined a custom feature in your type system (e.g., STRING suffixFeature; for the WORD type). Then use this rule to assign values:

// Match qualifying words and assign their mapped feature value
WORD{-> ASSIGN(suffixFeature, suffixMap.getValue())};

4. Expected Results for Your Input Text #

For your input:

yearbook book databook worship friendship yearbook

The rule will assign:

• b. to yearbook, book, databook (all end with book)
• sh. to worship, friendship (all end with ship)

Quick Tips #

• Verify your CSV file path: if it’s not in the same directory as your script, use the full path or place it in your project’s resource folder.
• If a word could match multiple suffixes (unlikely here), WORDTABLE prioritizes the longest matching suffix.
• Ensure your type system includes the custom feature you’re assigning (define this in the Ruta descriptor or a separate type system XML file).

内容的提问来源于stack exchange，提问作者Sugunalakshmi Pagemajik