This work evaluates the potential of large language models (LLMs) to power digital assistants capable of complex action execution. Such assistants rely on pre-trained programming knowledge to execute multi-step goals by composing objects and functions defined in assistant libraries into action execution programs. To achieve this, we develop ASPERA, a framework comprising an assistant library simulation and a human-assisted LLM data generation engine. Our engine allows developers to guide LLM generation of high-quality tasks consisting of complex user queries, simulation state and corresponding validation programs, tackling data availability and evaluation robustness challenges. Alongside the framework we release Asper-Bench, an evaluation dataset of 250 challenging tasks generated using ASPERA, which we use to show that program generation grounded in custom assistant libraries is a significant challenge to LLMs compared to dependency-free code generation.

Programmable task-oriented dialogue (TOD) agents enable language models to follow structured dialogue policies, but their effectiveness hinges on accurate dialogue state tracking (DST). We present PyTOD, an agent that generates executable code to track dialogue state and uses policy and execution feedback for efficient error correction. To achieve this, PyTOD employs a simple constrained decoding approach, using a language model instead of grammar rules to follow API schemata. This leads to state-of-the-art DST performance on the challenging SGD benchmark. Our experiments show that PyTOD surpasses strong baselines in both accuracy and cross-turn consistency, demonstrating the effectiveness of execution-aware state tracking.

We consider a new perspective on dialog state tracking (DST), the task of estimating a user’s goal through the course of a dialog. By formulating DST as a semantic parsing task over hierarchical representations, we can incorporate semantic compositionality, cross-domain knowledge sharing and co-reference. We present TreeDST, a dataset of 27k conversations annotated with tree-structured dialog states and system acts. We describe an encoder-decoder framework for DST with hierarchical representations, which leads to ~20% improvement over state-of-the-art DST approaches that operate on a flat meaning space of slot-value pairs.

We present our submission to the 2018 Duolingo Shared Task on Second Language Acquisition Modeling (SLAM). We focus on evaluating a range of features for the task, including user-derived measures, while examining how far we can get with a simple linear classifier. Our analysis reveals that errors differ per exercise format, which motivates our final and best-performing system: a task-wise (per exercise-format) model.

Multitask learning has been applied successfully to a range of tasks, mostly morphosyntactic. However, little is known on when MTL works and whether there are data characteristics that help to determine the success of MTL. In this paper we evaluate a range of semantic sequence labeling tasks in a MTL setup. We examine different auxiliary task configurations, amongst which a novel setup, and correlate their impact to data-dependent conditions. Our results show that MTL is not always effective, because significant improvements are obtained only for 1 out of 5 tasks. When successful, auxiliary tasks with compact and more uniform label distributions are preferable.

We present UDP, the first training-free parser for Universal Dependencies (UD). Our algorithm is based on PageRank and a small set of specific dependency head rules. UDP features two-step decoding to guarantee that function words are attached as leaf nodes. The parser requires no training, and it is competitive with a delexicalized transfer system. UDP offers a linguistically sound unsupervised alternative to cross-lingual parsing for UD. The parser has very few parameters and distinctly robust to domain change across languages.

The Conference on Computational Natural Language Learning (CoNLL) features a shared task, in which participants train and test their learning systems on the same data sets. In 2017, the task was devoted to learning dependency parsers for a large number of languages, in a real-world setting without any gold-standard annotation on input. All test sets followed a unified annotation scheme, namely that of Universal Dependencies. In this paper, we define the task and evaluation methodology, describe how the data sets were prepared, report and analyze the main results, and provide a brief categorization of the different approaches of the participating systems.

We focus on the identification of omission in statement pairs. We compare three annotation schemes, namely two different crowdsourcing schemes and manual expert annotation. We show that the simplest of the two crowdsourcing approaches yields a better annotation quality than the more complex one. We use a dedicated classifier to assess whether the annotators’ behavior can be explained by straightforward linguistic features. The classifier benefits from a modeling that uses lexical information beyond length and overlap measures. However, for our task, we argue that expert and not crowdsourcing-based annotation is the best compromise between annotation cost and quality.

This article evaluates the extension of a dependency parser that performs joint syntactic analysis and multiword expression identification. We show that, given sufficient training data, the parser benefits from explicit multiword information and improves overall labeled accuracy score in eight of the ten evaluation cases.

Neural part-of-speech tagging has achieved competitive results with the incorporation of character-based and pre-trained word embeddings. In this paper, we show that a state-of-the-art bi-LSTM tagger can benefit from using information from morphosyntactic lexicons as additional input. The tagger, trained on several dozen languages, shows a consistent, average improvement when using lexical information, even when also using character-based embeddings, thus showing the complementarity of the different sources of lexical information. The improvements are particularly important for the smaller datasets.

In this article we present an expansion of the supersense inventory. All new super-senses are extensions of members of the current inventory, which we postulate by identifying semantically coherent groups of synsets. We cover the expansion of the already-established supernsense inventory for nouns and verbs, the addition of coarse supersenses for adjectives in absence of a canonical supersense inventory, and super-senses for verbal satellites. We evaluate the viability of the new senses examining the annotation agreement, frequency and co-ocurrence patterns.

Approximate summary optimisation for selections of ROUGE It is standard to measure automatic summariser performance using the ROUGE metric. Unfortunately, ROUGE is not appropriate for unsupervised summarisation approaches. On the other hand, we show that it is possible to optimise approximately for ROUGE-n by using a document-weighted ROUGE objective. Doing so results in state-of-the-art summariser performance for single and multiple document summaries for both English and French. This is despite a non-correlation of the documentweighted ROUGE metric with human judgments, unlike the original ROUGE metric. These findings suggest a theoretical approximation link between the two metrics.

We launch the SemDaX corpus which is a recently completed Danish human-annotated corpus available through a CLARIN academic license. The corpus includes approx. 90,000 words, comprises six textual domains, and is annotated with sense inventories of different granularity. The aim of the developed corpus is twofold: i) to assess the reliability of the different sense annotation schemes for Danish measured by qualitative analyses and annotation agreement scores, and ii) to serve as training and test data for machine learning algorithms with the practical purpose of developing sense taggers for Danish. To these aims, we take a new approach to human-annotated corpus resources by double annotating a much larger part of the corpus than what is normally seen: for the all-words task we double annotated 60% of the material and for the lexical sample task 100%. We include in the corpus not only the adjucated files, but also the diverging annotations. In other words, we consider not all disagreement to be noise, but rather to contain valuable linguistic information that can help us improve our annotation schemes and our learning algorithms.

We propose a novel approach to cross-lingual part-of-speech tagging and dependency parsing for truly low-resource languages. Our annotation projection-based approach yields tagging and parsing models for over 100 languages. All that is needed are freely available parallel texts, and taggers and parsers for resource-rich languages. The empirical evaluation across 30 test languages shows that our method consistently provides top-level accuracies, close to established upper bounds, and outperforms several competitive baselines.

User-generated content presents many challenges for its automatic processing. While many of them do come from out-of-vocabulary effects, others spawn from different linguistic phenomena such as unusual syntax. In this work we present a French three-domain data set made up of question headlines from a cooking forum, game chat logs and associated forums from two popular online games (MINECRAFT & LEAGUE OF LEGENDS). We chose these domains because they encompass different degrees of lexical and syntactic compliance with canonical language. We conduct an automatic and manual evaluation of the difficulties of processing these domains for part-of-speech prediction, and introduce a pilot study to determine whether dependency analysis lends itself well to annotate these data. We also discuss the development cost of our data set.

This article outlines a methodology that uses crowdsourcing to reduce the workload of experts for complex semantic tasks. We split turker-annotated datasets into a high-agreement block, which is not modified, and a low-agreement block, which is re-annotated by experts. The resulting annotations have higher observed agreement. We identify different biases in the annotation for both turkers and experts.

The following work describes a voting system to automatically classify the sense selection of the complex types Location/Organization and Container/Content, which depend on regular polysemy, as described by the Generative Lexicon (Pustejovsky, 1995) . This kind of sense alternations very often presents semantic underspecificacion between its two possible selected senses. This kind of underspecification is not traditionally contemplated in word sense disambiguation systems, as disambiguation systems are still coping with the need of a representation and recognition of underspecification (Pustejovsky, 2009) The data are characterized by the morphosyntactic and lexical enviroment of the headwords and provided as input for a classifier. The baseline decision tree classifier is compared against an eight-member voting scheme obtained from variants of the training data generated by modifications on the class representation and from two different classification algorithms, namely decision trees and k-nearest neighbors. The voting system improves the accuracy for the non-underspecified senses, but the underspecified sense remains difficult to identify