Learned Sparse Retrieval (LSR) models use vocabularies from pre-trained transformers, which often split entities into nonsensical fragments. Splitting entities diminishes retrieval accuracy and limits the model’s ability to incorporate up-to-date world knowledge not included in the training data. In this work, we enhance the LSR vocabulary with Wikipedia concepts and entities, enabling the model to resolve ambiguities more effectively and stay current with evolving knowledge. Central to our approach is a Dynamic Vocabulary (DyVo) head, which leverages existing entity embeddings and an entity retrieval component that identifies entities relevant to a query or document. We use the DyVo head to generate entity weights, which are then merged with word piece weights to create joint representations for efficient indexing and retrieval using an inverted index. In experiments across three entity-rich document ranking datasets, the resulting DyVo model substantially outperforms several state-of-the-art baselines.

In recent years, research shows that neural ranking models (NRMs) substantially outperform their lexical counterparts in text retrieval. In traditional search pipelines, a combination of features leads to well-defined behaviour. However, as neural approaches become increasingly prevalent as the final scoring component of engines or as standalone systems, their robustness to malicious text and, more generally, semantic perturbation needs to be better understood. We posit that the transformer attention mechanism can induce exploitable defects in search models through sensitivity to token position within a sequence, leading to an attack that could generalise beyond a single query or topic. We demonstrate such defects by showing that non-relevant text–such as promotional content–can be easily injected into a document without adversely affecting its position in search results. Unlike previous gradient-based attacks, we demonstrate the existence of these biases in a query-agnostic fashion. In doing so, without the knowledge of topicality, we can still reduce the negative effects of non-relevant content injection by controlling injection position. Our experiments are conducted with simulated on-topic promotional text automatically generated by prompting LLMs with topical context from target documents. We find that contextualisation of a non-relevant text further reduces negative effects whilst likely circumventing existing content filtering mechanisms. In contrast, lexical models are found to be more resilient to such content injection attacks. We then investigate a simple yet effective compensation for the weaknesses of the NRMs in search, validating our hypotheses regarding transformer bias.

Verbatim queries submitted to search engines often do not sufficiently describe the user’s search intent. Pseudo-relevance feedback (PRF) techniques, which modify a query’srepresentation using the top-ranked documents, have been shown to overcome such inadequacies and improve retrieval effectiveness for both lexical methods (e.g., BM25) and dense methods (e.g., ANCE, ColBERT). For instance, the recent ColBERT-PRF approach heuristically chooses new embeddings to add to the query representation using the inverse document frequency (IDF) of the underlying tokens. However, this heuristic potentially ignores the valuable context encoded by the embeddings. In this work, we present a contrastive solution that learns to select the most useful embeddings for expansion. More specifically, a deep language model-based contrastive weighting model, called CWPRF, is trained to learn to discriminate between relevant and non-relevant documents for semantic search. Our experimental results show that our contrastive weighting model can aid to select useful expansion embeddings and outperform various baselines. In particular, CWPRF can improve nDCG@10 by upto to 4.1% compared to an existing PRF approach for ColBERT while maintaining its efficiency.

Social media are heavily used by many users to share their mental health concerns and diagnoses. This trend has turned social media into a large-scale resource for researchers focused on detecting mental health conditions. Social media usage varies considerably across individuals. Thus, classification of patterns, including detecting signs of depression, must account for such variation. We address the disparity in classification effectiveness for users with little activity (e.g., new users). Our evaluation, performed on a large-scale dataset, shows considerable detection discrepancy based on user posting frequency. For instance, the F1 detection score of users with an above-median versus below-median number of posts is greater than double (0.803 vs 0.365) using a conventional CNN-based model; similar results were observed on lexical and transformer-based classifiers. To complement this evaluation, we propose a dynamic thresholding technique that adjusts the classifier’s sensitivity as a function of the number of posts a user has. This technique alone reduces the margin between users with many and few posts, on average, by 45% across all methods and increases overall performance, on average, by 33%. These findings emphasize the importance of evaluating and tuning natural language systems for potentially vulnerable populations.

Pretrained contextualized language models such as BERT and T5 have established a new state-of-the-art for ad-hoc search. However, it is not yet well understood why these methods are so effective, what makes some variants more effective than others, and what pitfalls they may have. We present a new comprehensive framework for Analyzing the Behavior of Neural IR ModeLs (ABNIRML), which includes new types of diagnostic probes that allow us to test several characteristics—such as writing styles, factuality, sensitivity to paraphrasing and word order—that are not addressed by previous techniques. To demonstrate the value of the framework, we conduct an extensive empirical study that yields insights into the factors that contribute to the neural model’s gains, and identify potential unintended biases the models exhibit. Some of our results confirm conventional wisdom, for example, that recent neural ranking models rely less on exact term overlap with the query, and instead leverage richer linguistic information, evidenced by their higher sensitivity to word and sentence order. Other results are more surprising, such as that some models (e.g., T5 and ColBERT) are biased towards factually correct (rather than simply relevant) texts. Further, some characteristics vary even for the same base language model, and other characteristics can appear due to random variations during model training.1

Despite the recent successes of transformer-based models in terms of effectiveness on a variety of tasks, their decisions often remain opaque to humans. Explanations are particularly important for tasks like offensive language or toxicity detection on social media because a manual appeal process is often in place to dispute automatically flagged content. In this work, we propose a technique to improve the interpretability of these models, based on a simple and powerful assumption: a post is at least as toxic as its most toxic span. We incorporate this assumption into transformer models by scoring a post based on the maximum toxicity of its spans and augmenting the training process to identify correct spans. We find this approach effective and can produce explanations that exceed the quality of those provided by Logistic Regression analysis (often regarded as a highly-interpretable model), according to a human study.

Progress on NLP for mental health — indeed, for healthcare in general — is hampered by obstacles to shared, community-level access to relevant data. We report on what is, to our knowledge, the first attempt to address this problem in mental health by conducting a shared task using sensitive data in a secure data enclave. Participating teams received access to Twitter posts donated for research, including data from users with and without suicide attempts, and did all work with the dataset entirely within a secure computational environment. We discuss the task, team results, and lessons learned to set the stage for future tasks on sensitive or confidential data.

Offensive language detection is an important and challenging task in natural language processing. We present our submissions to the OffensEval 2020 shared task, which includes three English sub-tasks: identifying the presence of offensive language (Sub-task A), identifying the presence of target in offensive language (Sub-task B), and identifying the categories of the target (Sub-task C). Our experiments explore using a domain-tuned contextualized language model (namely, BERT) for this task. We also experiment with different components and configurations (e.g., a multi-view SVM) stacked upon BERT models for specific sub-tasks. Our submissions achieve F1 scores of 91.7% in Sub-task A, 66.5% in Sub-task B, and 63.2% in Sub-task C. We perform an ablation study which reveals that domain tuning considerably improves the classification performance. Furthermore, error analysis shows common misclassification errors made by our model and outlines research directions for future.

With worldwide concerns surrounding the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), there is a rapidly growing body of scientific literature on the virus. Clinicians, researchers, and policy-makers need to be able to search these articles effectively. In this work, we present a zero-shot ranking algorithm that adapts to COVID-related scientific literature. Our approach filters training data from another collection down to medical-related queries, uses a neural re-ranking model pre-trained on scientific text (SciBERT), and filters the target document collection. This approach ranks top among zero-shot methods on the TREC COVID Round 1 leaderboard, and exhibits a P@5 of 0.80 and an nDCG@10 of 0.68 when evaluated on both Round 1 and 2 judgments. Despite not relying on TREC-COVID data, our method outperforms models that do. As one of the first search methods to thoroughly evaluate COVID-19 search, we hope that this serves as a strong baseline and helps in the global crisis.

We investigate the effect of various dependency-based word embeddings on distinguishing between functional and domain similarity, word similarity rankings, and two downstream tasks in English. Variations include word embeddings trained using context windows from Stanford and Universal dependencies at several levels of enhancement (ranging from unlabeled, to Enhanced++ dependencies). Results are compared to basic linear contexts and evaluated on several datasets. We found that embeddings trained with Universal and Stanford dependency contexts excel at different tasks, and that enhanced dependencies often improve performance.

Mental health is a significant and growing public health concern. As language usage can be leveraged to obtain crucial insights into mental health conditions, there is a need for large-scale, labeled, mental health-related datasets of users who have been diagnosed with one or more of such conditions. In this paper, we investigate the creation of high-precision patterns to identify self-reported diagnoses of nine different mental health conditions, and obtain high-quality labeled data without the need for manual labelling. We introduce the SMHD (Self-reported Mental Health Diagnoses) dataset and make it available. SMHD is a novel large dataset of social media posts from users with one or multiple mental health conditions along with matched control users. We examine distinctions in users’ language, as measured by linguistic and psychological variables. We further explore text classification methods to identify individuals with mental conditions through their language.

Self-reported diagnosis statements have been widely employed in studying language related to mental health in social media. However, existing research has largely ignored the temporality of mental health diagnoses. In this work, we introduce RSDD-Time: a new dataset of 598 manually annotated self-reported depression diagnosis posts from Reddit that include temporal information about the diagnosis. Annotations include whether a mental health condition is present and how recently the diagnosis happened. Furthermore, we include exact temporal spans that relate to the date of diagnosis. This information is valuable for various computational methods to examine mental health through social media because one’s mental health state is not static. We also test several baseline classification and extraction approaches, which suggest that extracting temporal information from self-reported diagnosis statements is challenging.

SemEval 2018 Task 7 focuses on relation extraction and classification in scientific literature. In this work, we present our tree-based LSTM network for this shared task. Our approach placed 9th (of 28) for subtask 1.1 (relation classification), and 5th (of 20) for subtask 1.2 (relation classification with noisy entities). We also provide an ablation study of features included as input to the network.

Clinical TempEval 2017 (SemEval 2017 Task 12) addresses the task of cross-domain temporal extraction from clinical text. We present a system for this task that uses supervised learning for the extraction of temporal expression and event spans with corresponding attributes and narrative container relations. Approaches include conditional random fields and decision tree ensembles, using lexical, syntactic, semantic, distributional, and rule-based features. Our system received best or second best scores in TIMEX3 span, EVENT span, and CONTAINS relation extraction.