We introduce GENIUS: a conditional text generation model using sketches as input, which can fill in the missing contexts for a given sketch (key information consisting of textual spans, phrases, or words, concatenated by mask tokens). GENIUS is pre-trained on a large-scale textual corpus with a novel reconstruction from sketch objective using an extreme and selective masking strategy, enabling it to generate diverse and high-quality texts given sketches. Comparison with other competitive conditional language models (CLMs) reveals the superiority of GENIUS's text generation quality. We further show that GENIUS can be used as a strong and ready-to-use data augmentation tool for various natural language processing (NLP) tasks. Most existing textual data augmentation methods are either too conservative, by making small changes to the original text, or too aggressive, by creating entirely new samples. With GENIUS, we propose GeniusAug, which first extracts the target-aware sketches from the original training set and then generates new samples based on the sketches. Empirical experiments on 6 text classification datasets show that GeniusAug significantly improves the models' performance in both in-distribution (ID) and out-of-distribution (OOD) settings. We also demonstrate the effectiveness of GeniusAug on named entity recognition (NER) and machine reading comprehension (MRC) tasks. (Code and models are publicly available at https://github.com/microsoft/SCGLab and https://github.com/beyondguo/genius)

The question of answering queries over ML predictions has been gaining attention in the database community. This question is challenging because the cost of finding high quality answers corresponds to invoking an oracle such as a human expert or an expensive deep neural network model on every single item in the DB and then applying the query. We develop a novel unified framework for approximate query answering by leveraging a proxy to minimize the oracle usage of finding high quality answers for both Precision-Target (PT) and Recall-Target (RT) queries. Our framework uses a judicious combination of invoking the expensive oracle on data samples and applying the cheap proxy on the objects in the DB. It relies on two assumptions. Under the Proxy Quality assumption, proxy quality can be quantified in a probabilistic manner w.r.t. the oracle. This allows us to develop two algorithms: PQA that efficiently finds high quality answers with high probability and no oracle calls, and PQE, a heuristic extension that achieves empirically good performance with a small number of oracle calls. Alternatively, under the Core Set Closure assumption, we develop two algorithms: CSC that efficiently returns high quality answers with high probability and minimal oracle usage, and CSE, which extends it to more general settings. Our extensive experiments on five real-world datasets on both query types, PT and RT, demonstrate that our algorithms outperform the state-of-the-art and achieve high result quality with provable statistical guarantees.

To provide click simulation or relevance estimation based on users' implicit interaction feedback, click models have been much studied during recent years. Most click models focus on user behaviors towards a single list. However, with the development of user interface (UI) design, the layout of displayed items on a result page tends to be multi-block (i.e., multi-list) style instead of a single list, which requires different assumptions to model user behaviors more accurately. There exist click models for multi-block pages in desktop contexts, but they cannot be directly applied to mobile scenarios due to different interaction manners, result types and especially multi-block presentation styles. In particular, multi-block mobile pages can normally be decomposed into interleavings of basic vertical blocks and horizontal blocks, thus resulting in typically F-shape forms. To mitigate gaps between desktop and mobile contexts for multi-block pages, we conduct a user eye-tracking study, and identify users' sequential browsing, block skip and comparison patterns on F-shape pages. These findings lead to the design of a novel F-shape Click Model (FSCM), which serves as a general solution to multi-block mobile pages. Firstly, we construct a directed acyclic graph (DAG) for each page, where each item is regarded as a vertex and each edge indicates the user's possible examination flow. Secondly, we propose DAG-structured GRUs and a comparison module to model users' sequential (sequential browsing, block skip) and non-sequential (comparison) behaviors respectively. Finally, we combine GRU states and comparison patterns to perform user click predictions. Experiments on a large-scale real-world dataset validate the effectiveness of FSCM on user behavior predictions compared with baseline models.

Clinical Cohort Studies (CCS), such as randomized clinical trials, are a great source of documented clinical research. Ideally, a clinical expert inspects these articles for exploratory analysis ranging from drug discovery for evaluating the efficacy of existing drugs in tackling emerging diseases to the first test of newly developed drugs. However, more than 100 articles are published daily on a single prevalent disease like COVID-19 in PubMed. As a result, it can take days for a physician to find articles and extract relevant information. Can we develop a system to sift through the long list of these articles faster and document the crucial takeaways from each of these articles? In this work, we propose CCS Explorer, an end-to-end system for relevance prediction of sentences, extractive summarization, and patient, outcome, and intervention entity detection from CCS. CCS Explorer is packaged in a web-based graphical user interface where the user can provide any disease name. CCS Explorer then extracts and aggregates all relevant information from articles on PubMed based on the results of an automatically generated query produced on the back-end. For each task, CCS Explorer fine-tunes pre-trained language representation models based on transformers with additional layers. The models are evaluated using two publicly available datasets. CCS Explorer obtains a recall of 80.2%, AUC-ROC of 0.843, and an accuracy of 88.3% on sentence relevance prediction using BioBERT and achieves an average Micro F1-Score of 77.8% on Patient, Intervention, Outcome detection (PIO) using PubMedBERT. Thus, CCS Explorer can reliably extract relevant information to summarize articles, saving time by $\sim \text{660}\times$.

For humans and computers, the first step in answering an open-domain question is retrieving a set of relevant documents from a large corpus. However, the strategies that computers use fundamentally differ from those of humans. To better understand these differences, we design a gamified interface for data collection -- Cheater's Bowl -- where a human answers complex questions with access to both traditional and modern search tools. We collect a dataset of human search sessions, analyze human search strategies, and compare them to state-of-the-art multi-hop QA models. Humans query logically, apply dynamic search chains, and use world knowledge to boost searching. We demonstrate how human queries can improve the accuracy of existing systems and propose improving the future design of QA models.

Question answering over knowledge bases (KBQA) has become a popular approach to help users extract information from knowledge bases. Although several systems exist, choosing one suitable for a particular application scenario is difficult. In this article, we provide a comparative study of six representative KBQA systems on eight benchmark datasets. In that, we study various question types, properties, languages, and domains to provide insights on where existing systems struggle. On top of that, we propose an advanced mapping algorithm to aid existing models in achieving superior results. Moreover, we also develop a multilingual corpus COVID-KGQA, which encourages COVID-19 research and multilingualism for the diversity of future AI. Finally, we discuss the key findings and their implications as well as performance guidelines and some future improvements. Our source code is available at \url{https://github.com/tamlhp/kbqa}.

Keyword extraction has been an important topic for modern natural language processing. With its applications ranging from ontology generation, fact verification in summarized text, and recommendation systems. While it has had significant data-intensive applications, it is often hampered when the data set is small. Downstream training for keyword extractors is a lengthy process and requires a significant amount of data. Recently, Few-shot Learning (FSL) and Zero-Shot Learning (ZSL) have been proposed to tackle this problem. Therefore, we propose AdaptKeyBERT, a pipeline for training keyword extractors with LLM bases by incorporating the concept of regularized attention into a pre-training phase for downstream domain adaptation.

In this thesis, we investigated the relevance, faithfulness, and succinctness aspects of Long Form Question Answering (LFQA). LFQA aims to generate an in-depth, paragraph-length answer for a given question, to help bridge the gap between real scenarios and the existing open-domain QA models which can only extract short-span answers. LFQA is quite challenging and under-explored. Few works have been done to build an effective LFQA system. It is even more challenging to generate a good-quality long-form answer relevant to the query and faithful to facts, since a considerable amount of redundant, complementary, or contradictory information will be contained in the retrieved documents. Moreover, no prior work has been investigated to generate succinct answers. We are among the first to research the LFQA task. We pioneered the research direction to improve the answer quality in terms of 1) query-relevance, 2) answer faithfulness, and 3) answer succinctness.

In this paper, we present DuReader_retrieval, a large-scale Chinese dataset for passage retrieval. DuReader_retrieval contains more than 90K queries and over 8M unique passages from a commercial search engine. To alleviate the shortcomings of other datasets and ensure the quality of our benchmark, we (1) reduce the false negatives in development and test sets by manually annotating results pooled from multiple retrievers, and (2) remove the training queries that are semantically similar to the development and testing queries. Additionally, we provide two out-of-domain testing sets for cross-domain evaluation, as well as a set of human translated queries for for cross-lingual retrieval evaluation. The experiments demonstrate that DuReader_retrieval is challenging and a number of problems remain unsolved, such as the salient phrase mismatch and the syntactic mismatch between queries and paragraphs. These experiments also show that dense retrievers do not generalize well across domains, and cross-lingual retrieval is essentially challenging. DuReader_retrieval is publicly available at https://github.com/baidu/DuReader/tree/master/DuReader-Retrieval.

Recent progress in language model pre-training has led to important improvements in Named Entity Recognition (NER). Nonetheless, this progress has been mainly tested in well-formatted documents such as news, Wikipedia, or scientific articles. In social media the landscape is different, in which it adds another layer of complexity due to its noisy and dynamic nature. In this paper, we focus on NER in Twitter, one of the largest social media platforms, and construct a new NER dataset, TweetNER7, which contains seven entity types annotated over 11,382 tweets from September 2019 to August 2021. The dataset was constructed by carefully distributing the tweets over time and taking representative trends as a basis. Along with the dataset, we provide a set of language model baselines and perform an analysis on the language model performance on the task, especially analyzing the impact of different time periods. In particular, we focus on three important temporal aspects in our analysis: short-term degradation of NER models over time, strategies to fine-tune a language model over different periods, and self-labeling as an alternative to lack of recently-labeled data. TweetNER7 is released publicly (https://huggingface.co/datasets/tner/tweetner7) along with the models fine-tuned on it.