Large pre-trained language models for textual data have an unconstrained
output space; at each decoding step, they can produce any of 10,000s of
sub-word tokens. When fine-tuned to target constrained formal languages like
SQL, these models often generate invalid code, rendering it unusable. We
propose PICARD (code and trained models available at
this https URL), a method for constraining auto-regressive
decoders of language models through incremental parsing. PICARD helps to find
valid output sequences by rejecting inadmissible tokens at each decoding step.
On the challenging Spider and CoSQL text-to-SQL translation tasks, we show that
PICARD transforms fine-tuned T5 models with passable performance into
state-of-the-art solutions.

本研究提出了一种名为 PICARD 的方法，它通过增量解析来约束自回归语言模型解码器，并在需要的时候拒绝不合法的令牌。在 Spider 和 CoSQL 的文本到 SQL 翻译任务中，我们发现 PICARD 将微调后的 T5 模型转化为业界领先的解决方案。

PICARD: 从语言模型中进行约束的自回归解码的增量解析

PICARD: Parsing Incrementally for Constrained Auto-Regressive Decoding  from Language Models

Complex Word Identification (CWI) is a task centered on detecting
hard-to-understand words, or groups of words, in texts from different areas of
expertise. The purpose of CWI is to highlight problematic structures that
non-native speakers would usually find difficult to understand. Our approach
uses zero-shot, one-shot, and few-shot learning techniques, alongside
state-of-the-art solutions for Natural Language Processing (NLP) tasks (i.e.,
Transformers). Our aim is to provide evidence that the proposed models can
learn the characteristics of complex words in a multilingual environment by
relying on the CWI shared task 2018 dataset available for four different
languages (i.e., English, German, Spanish, and also French). Our approach
surpasses state-of-the-art cross-lingual results in terms of macro F1-score on
English (0.774), German (0.782), and Spanish (0.734) languages, for the
zero-shot learning scenario. At the same time, our model also outperforms the
state-of-the-art monolingual result for German (0.795 macro F1-score).

本研究利用零样本、一次样本和少量样本学习技术，结合自然语言处理的最先进解决方案，通过使用四种不同语言（英语、德语、西班牙语和法语）的 CWI 共享任务数据集，证明了提出的模型可以在多语言环境中学习复杂单词的特征，并在零样本学习场景下在英语、德语和西班牙语三种语言中的宏 F1 分数上超过了现有的跨语言结果。同时，我们的模型也在德语（0.795 宏 F1 分数）的单语上表现出色。

跨语言迁移学习用于复杂词汇识别

Cross-Lingual Transfer Learning for Complex Word Identification

Answer Set Programming (ASP) is a well-established paradigm of declarative
programming in close relationship with other declarative formalisms such as SAT
Modulo Theories, Constraint Handling Rules, FO(.), PDDL and many others. Since
its first informal editions, ASP systems have been compared in the now
well-established ASP Competition. The Third (Open) ASP Competition, as the
sequel to the ASP Competitions Series held at the University of Potsdam in
Germany (2006-2007) and at the University of Leuven in Belgium in 2009, took
place at the University of Calabria (Italy) in the first half of 2011.
Participants competed on a pre-selected collection of benchmark problems, taken
from a variety of domains as well as real world applications. The Competition
ran on two tracks: the Model and Solve (M&S) Track, based on an open problem
encoding, and open language, and open to any kind of system based on a
declarative specification paradigm; and the System Track, run on the basis of
fixed, public problem encodings, written in a standard ASP language. This paper
discusses the format of the Competition and the rationale behind it, then
reports the results for both tracks. Comparison with the second ASP competition
and state-of-the-art solutions for some of the benchmark domains is eventually
discussed.
To appear in Theory and Practice of Logic Programming (TPLP).

本文报道了第三届 (开放式) ASP 竞赛的格式以及结果。该竞赛旨在比较不同基于声明性编程的系统在一系列基准问题上的表现，并讨论了竞赛与第二届 ASP 竞赛以及一些基准领域的最新解决方案的比较。