<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Structured Rationales | Language Technologies Lab</title><link>http://nlp.unibo.it/tag/structured-rationales/</link><atom:link href="http://nlp.unibo.it/tag/structured-rationales/index.xml" rel="self" type="application/rss+xml"/><description>Structured Rationales</description><generator>Hugo Blox Builder (https://hugoblox.com)</generator><language>en-us</language><lastBuildDate>Mon, 02 Mar 2026 00:00:00 +0000</lastBuildDate><image><url>http://nlp.unibo.it/media/icon_hu_7613a4a452ac7087.png</url><title>Structured Rationales</title><link>http://nlp.unibo.it/tag/structured-rationales/</link></image><item><title>Structured Rationalization via Tree kernel methods</title><link>http://nlp.unibo.it/proposals_interpretability/treekernels/</link><pubDate>Mon, 02 Mar 2026 00:00:00 +0000</pubDate><guid>http://nlp.unibo.it/proposals_interpretability/treekernels/</guid><description>&lt;p>&lt;strong>Description:&lt;/strong>&lt;br>
There are several techniques for transforming text into abstract structured content (AMR graphs, Parse trees, etc&amp;hellip;).
We are interested in applying rationalization in these contexts by also enforcing some structural constraints depending on the given scenario of application.
The constraints describe which type of allowed structured the rationalization system can extract.
In the case of tree kernels, these structures are different types of trees.&lt;/p>
&lt;p>&lt;strong>Contact:&lt;/strong> &lt;a href="mailto:federico.ruggeri6@unibo.it">Federico Ruggeri&lt;/a>&lt;/p>
&lt;p>&lt;strong>References:&lt;/strong>&lt;/p>
&lt;p>&lt;strong>Tree-constrained Graph Neural Networks for Argument Mining&lt;/strong>&lt;br>
Federico Ruggeri, Marco Lippi, Paolo Torroni&lt;br>
September 2021&lt;br>
&lt;a href="https://arxiv.org/abs/2110.00124" target="_blank" rel="noopener">PDF&lt;/a>&lt;/p></description></item></channel></rss>