The predictions of question answering (QA) systems are typically evaluated against manually annotated finite sets of one or more answers. This leads to a coverage limitation that results in underestimating the true performance of systems, and is typically addressed by extending over exact match (EM) with predefined rules or with the token-level F1 measure.In this paper, we present the first systematic conceptual and data-driven analysis to examine the shortcomings of token-level equivalence measures.To this end, we define the asymmetric notion of answer equivalence (AE), accepting answers that are equivalent to or improve over the reference, and publish over 23k human judgements for candidates produced by multiple QA systems on SQuAD.Through a careful analysis of this data, we reveal and quantify several concrete limitations of the F1 measure, such as a false impression of graduality, or missing dependence on the question.Since collecting AE annotations for each evaluated model is expensive, we learn a BERT matching (BEM) measure to approximate this task. Being a simpler task than QA, we find BEM to provide significantly better AE approximations than F1, and to more accurately reflect the performance of systems.Finally, we demonstrate the practical utility of AE and BEM on the concrete application of minimal accurate prediction sets, reducing the number of required answers by up to X2.6.
We release FoolMeTwice (FM2 for short), a large dataset of challenging entailment pairs collected through a fun multi-player game. Gamification encourages adversarial examples, drastically lowering the number of examples that can be solved using “shortcuts” compared to other popular entailment datasets. Players are presented with two tasks. The first task asks the player to write a plausible claim based on the evidence from a Wikipedia page. The second one shows two plausible claims written by other players, one of which is false, and the goal is to identify it before the time runs out. Players “pay” to see clues retrieved from the evidence pool: the more evidence the player needs, the harder the claim. Game-play between motivated players leads to diverse strategies for crafting claims, such as temporal inference and diverting to unrelated evidence, and results in higher quality data for the entailment and evidence retrieval tasks. We open source the dataset and the game code.
In addition to the traditional task of machines answering questions, question answering (QA) research creates interesting, challenging questions that help systems how to answer questions and reveal the best systems. We argue that creating a QA dataset—and the ubiquitous leaderboard that goes with it—closely resembles running a trivia tournament: you write questions, have agents (either humans or machines) answer the questions, and declare a winner. However, the research community has ignored the hard-learned lessons from decades of the trivia community creating vibrant, fair, and effective question answering competitions. After detailing problems with existing QA datasets, we outline the key lessons—removing ambiguity, discriminating skill, and adjudicating disputes—that can transfer to QA research and how they might be implemented.
Stress has long been established as a major cue in word segmentation for English infants. We show that enabling a current state-of-the-art Bayesian word segmentation model to take advantage of stress cues noticeably improves its performance. We find that the improvements range from 10 to 4%, depending on both the use of phonotactic cues and, to a lesser extent, the amount of evidence available to the learner. We also find that in particular early on, stress cues are much more useful for our model than phonotactic cues by themselves, consistent with the finding that children do seem to use stress cues before they use phonotactic cues. Finally, we study how the model’s knowledge about stress patterns evolves over time. We not only find that our model correctly acquires the most frequent patterns relatively quickly but also that the Unique Stress Constraint that is at the heart of a previously proposed model does not need to be built in but can be acquired jointly with word segmentation.
This paper describes a multi-lingual large-scale concept network obtained automatically by mining for concepts and relations and exploiting a variety of sources of knowledge from Wikipedia. Concepts and their lexicalizations are extracted from Wikipedia pages, in particular from article titles, hyperlinks, disambiguation pages and cross-language links. Relations are extracted from the category and page network, from the category names, from infoboxes and the body of the articles. The resulting network has two main components: (i) a central, language independent index of concepts, which serves to keep track of the concepts' lexicalizations both within a language and across languages, and to separate linguistic expressions of concepts from the relations in which they are involved (concepts themselves are represented as numeric IDs); (ii) a large network built on the basis of the relations extracted, represented as relations between concepts (more specifically, the numeric IDs). The various stages of obtaining the network were separately evaluated, and the results show a qualitative resource.