Lillian Lee

3 paper cards in Mova's library · 2003–2021

Lillian Lee wrote 3 of the papers Mova reads, in 2003–2021, most of them on Dependency parsing, Part-of-speech tagging, Machine translation; often with Tianze Shi, Regina Barzilay.

The profile is the library's match of the name; two people with one name are not always told apart.

Ideas Mova took (1)

From the ideas ledger: an idea tried, kept or queued, or measured with no gain, and what it gave.

tried — Align all English renderings of one verse across the translations in the same way. The shared backbone is the frame-preserving part of the verse, and the slots are the author delta. A slot filled in three or more translations goes to the standard; a slot… (Learning to Paraphrase: An Unsupervised Approach Using… · translate r6-lattice: the four Bibles' renderings of each training verse aligned (LCS of concepts); the standard learned on the shared backbone only (1.20 M…)

Transition-based Bubble Parsing: Improvements on Coordination Structure Prediction (2021)
Standard dependency trees give coordination an awkward shape: one conjunct must be the head, and a shared modifier ("hot" in "hot coffee or tea and a bun") cannot be told apart from a private one.
Extracting Headless MWEs from Dependency Parse Trees: Parsing, Tagging, and Joint Modeling Approaches (2020)
A headless multi-word expression, such as "Wells Fargo" or "day after day", has no internal head, yet dependency schemes force a head on it.
Learning to Paraphrase: An Unsupervised Approach Using Multiple-Sequence Alignment (2003)
The paper addresses sentence-level paraphrase, which is harder than word or phrase paraphrase, using comparable news corpora: different sources reporting the same events.