We present two new large-scale datasets aimed at evaluating systems designed to comprehend a natural language query and extract its answer from a large corpus of text. The QUASAR-S dataset consists of 37000 cloze-style (fill-in-the-gap) queries constructed from definitions of software entity tags on the popular website Stack Overflow. The posts and comments on the website serve as the backgroun...

Visual Question Answering (VQA) aims to answer the natural language question about a given image by understanding multimodal content. However, quality of most existing visual-language pre-training (VLP) methods is still limited, mainly due to: (1) Incompatibility. Upstream tasks are generally incompatible with downstream answering tasks, which makes knowledge from model not well transferable an...

We used ERPs to investigate the pre-activation of form and meaning in language comprehension. Participants read high-cloze sentence contexts (e.g., ‘‘The student is going to the library to borrow a. . .”), followed by a word that was predictable (book), form-related (hook) or semantically related (page) to the predictable word, or unrelated (sofa). At a 500 ms SOA (Experiment 1), semantically r...

Representation and learning of commonsense knowledge is one of the foundational problems in the quest to enable deep language understanding. This issue is particularly challenging for understanding casual and correlational relationships between events. While this topic has received a lot of interest in the NLP community, research has been hindered by the lack of a proper evaluation framework. T...

Semantic unification, the process by which small blocks of semantic information are combined into a coherent utterance, has been studied with various types of tasks. However, whether the brain activations reported in these studies are attributed to semantic unification per se or to other task-induced concomitant processes still remains unclear. The neural basis for semantic unification in sente...

Schemata are expressions that are fixed except for slots available for novel words (I'm not a ______ person). Our goals were to quantify speakers' knowledge, examine semantic flexibility in open slots, and compare performance data in two generations of speakers using cloze procedures in formulaic expressions, schemata open slots, fixed portions of schemata, and novel sentences. Fewer unique wor...

The present research examined the effects of lexical cohesion cloze reading exercises (in which deletions involved various types cohesive devices) on improving EFL students’ comprehension. Sixty-four Thai university students were randomly assigned to control and experimental groups. Both groups received 12 sessions skill-based instruction, but group had additional practice in recognizing relati...