In recent years, large language models (LLMs) have demonstrated outstanding capabilities in various tasks. However, LLMs also have various drawbacks, especially hallucination. Hallucination refers to the generation of content that does not align with the user input, contradicts previously generated content or world knowledge. Current research on hallucination mainly include knowledge retrieval, prompt engineering, training data improvement, reinforcement learning, etc. However, these methods do not involve different categories of hallucinations which is important on hallucination analysis, and make detailed investigation for the internal state of LLMs which indicates the direction on hallucination occurrence. Therefore, in our research, we introduce an attribution framework to trace the origins of hallucinations based on the internal signals of LLMs. To support this framework, we develop a new benchmark named RelQA-Cate, which includes eight categories of hallucinations for the answers generated by LLMs. After that, we present a novel Differential Penalty Decoding (DPD) strategy for reducing hallucinations through adjusting post-probabilities of each answer. We conduct a series of experiments and the performance on answer reliability has significant improvement, achieving 28.25% at most, which demonstrates the effectiveness of our proposed DPD and its generalization in mitigating hallucination in LLMs.

The use of children's drawings to examining their conceptual understanding has been proven to be an effective method, but there are two major problems with previous research: 1. The content of the drawings heavily relies on the task, and the ecological validity of the conclusions is low; 2. The interpretation of drawings relies too much on the subjective feelings of the researchers. To address this issue, this study uses the Large Language Model (LLM) to identify 1420 children's scientific drawings (covering 9 scientific themes/concepts), and uses the word2vec algorithm to calculate their semantic similarity. The study explores whether there are consistent drawing representations for children on the same theme, and attempts to establish a norm for children's scientific drawings, providing a baseline reference for follow-up children's drawing research. The results show that the representation of most drawings has consistency, manifested as most semantic similarity>0.8. At the same time, it was found that the consistency of the representation is independent of the accuracy (of LLM’s recognition), indicating the existence of consistency bias. In the subsequent exploration of influencing factors, we used Kendall rank correlation coefficient to investigate the effects of "sample size", "abstract degree", and "focus points" on drawings, and used word frequency statistics to explore whether children represented abstract themes/concepts by reproducing what was taught in class. It was found that accuracy (of LLM’s recognition) is the most sensitive indicator, and data such as sample size and semantic similarity are related to it; The consistency between classroom experiments and teaching purpose is also an important factor, many students focus more on the experiments themselves rather than what they explain. In addition, most children tend to use examples they have seen in class to represent more abstract themes/concepts, indicating that they may need concrete examples to understand abstract things.

Our greatest source of insight into the real world today is via social media. Here, a major statement or quote by a public figure (world leader, politician, celebrity, scientist) can have wide-ranging impact, igniting extensive discussions and triggering reactions. It would be helpful to have tools for monitoring, querying, and inspecting the "flow" of social discourse. We introduce QuoteInspector, a system uniquely designed for efficient tracking and analysis of social media discussions around quotes. QuoteInspector leverages modern text embeddings and employs a clustering-based methodology for extracting topics from posts; it further integrates various NLP techniques for in-depth cluster analysis. Additionally, the system enhances the user experience by combining keyword- and relationship-based (structured) search for efficient and precise quote retrieval.

Retrieving relevant tables containing the necessary information to accurately answer a given question over tables is critical to open-domain question-answering (QA) systems. Previous methods assume the answer to such a question can be found either in a single table or multiple tables identified through question decomposition or rewriting. However, neither of these approaches is sufficient, as many questions require retrieving multiple tables and joining them through a join plan that cannot be discerned from the user query itself. If the join plan is not considered in the retrieval stage, the subsequent steps of reasoning and answering based on those retrieved tables are likely to be incorrect. To address this problem, we introduce a method that uncovers useful join relations for any query and database during table retrieval. We use a novel re-ranking method formulated as a mixed-integer program that considers not only table-query relevance but also table-table relevance that requires inferring join relationships. Our method outperforms the state-of-the-art approaches for table retrieval by up to 9.3% in F1 score and for end-to-end QA by up to 5.4% in accuracy.

As the trace norm can discover low-rank structures in a matrix, it has been widely used in multi-task learning to recover the low-rank structure contained in the parameter matrix. Recently, with the emerging of big complex datasets and the popularity of deep learning techniques, tensor trace norms have been used for deep multi-task models. However, existing tensor trace norms exhibit some limitations. For example, they cannot discover all the low-rank structures in a tensor, they require users to manually specify the importance of each component in the corresponding tensor trace norm, and they only capture the linear low-rank structure. To solve the first issue, in this paper, we propose a Generalized Tensor Trace Norm (GTTN). The GTTN is defined as a convex combination of matrix trace norms of all possible tensor flattenings and hence it can discover all the possible low-rank structures. For the second issue, in the induced objective function with the GTTN, we propose four strategies to learn combination coefficients in the GTTN. Furthermore, we propose the Nonlinear GTTN (NGTTN) to capture nonlinear low-rank structure among all the tasks. Experiments on benchmark datasets demonstrate the effectiveness of the proposed GTTN and NGTTN.

Many users turn to document retrieval systems (e.g. search engines) to seek answers to controversial questions. Answering such user queries usually require identifying responses within web documents, and aggregating the responses based on their different perspectives. Classical document retrieval systems fall short at delivering a set of direct and diverse responses to the users. Naturally, identifying such responses within a document is a natural language understanding task. In this paper, we examine the challenges of synthesizing such language understanding objectives with document retrieval, and study a new perspective-oriented document retrieval paradigm. We discuss and assess the inherent natural language understanding challenges in order to achieve the goal. Following the design challenges and principles, we demonstrate and evaluate a practical prototype pipeline system. We use the prototype system to conduct a user survey in order to assess the utility of our paradigm, as well as understanding the user information needs for controversial queries.

The trace norm is widely used in multi-task learning as it can discover low-rank structures among tasks in terms of model parameters. Nowadays, with the emerging of big complex datasets and the popularity of deep learning techniques, tensor trace norms have been used for deep multi-task models. However, existing tensor trace norms cannot discover all the low-rank structures and they require users to determine the importance of their components manually. To solve those two issues, in this paper, we propose a Generalized Tensor Trace Norm (GTTN). The GTTN is defined as a convex combination of matrix trace norms of all possible tensor flattenings and hence it can discover all the possible low-rank structures. Based on the induced objective function with the GTTN, we can learn combination coefficients in the GTTN with several strategies. Experiments on real-world datasets demonstrate the effectiveness of the proposed GTTN.

Yi Zhang, Zachary Ives, Dan Roth. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.

People rely on digital task management tools, such as email or to-do apps, to manage their tasks. Some of these tasks are large and complex, leading to action paralysis and feelings of being overwhelmed on the part of the user. The micro-productivity literature has shown that such tasks could benefit from being decomposed and organized, in order to reduce user cognitive load. Thus in this paper, we propose a novel end-to-end pipeline that consumes a complex task and induces a dependency graph from unstructured text to represent sub-tasks and their relationships. Our solution first finds nodes for sub-tasks from multiple 'how-to' articles on the web by injecting a neural text generator with three key desiderata - relevance, abstraction, and consensus. Then we resolve and infer edges between these subtask nodes by learning task dependency relations. We collect a new dataset of complex tasks with their sub-task graph to develop and evaluate our solutions. Both components of our graph induction solution are evaluated in experiments, demonstrating that our models outperform a state-of-the-art text generator significantly. Our generalizable and scalable end-to-end solution has important implications for boosting user productivity and assisting with digital task management.

In an era where generating content and publishing it is so easy, we are bombarded with information and are exposed to all kinds of claims, some of which do not always rank high on the truth scale. This paper suggests that the key to a longer-term, holistic, and systematic approach to navigating this information pollution is capturing the provenance of claims. To do that, we develop a formal definition of provenance graph for a given natural language claim, aiming to understand where the claim may come from and how it has evolved. To construct the graph, we model provenance inference, formulated mainly as an information extraction task and addressed via a textual entailment model. We evaluate our approach using two benchmark datasets, showing initial success in capturing the notion of provenance and its effectiveness on the application of claim verification.

Many modern data science applications build on data lakes, schema-agnostic repositories of data files and data products that offer limited organization and management capabilities. There is a need to build data lake search capabilities into data science environments, so scientists and analysts can find tables, schemas, workflows, and datasets useful to their task at hand. We develop search and management solutions for the Jupyter Notebook data science platform, to enable scientists to augment training data, find potential features to extract, clean data, and find joinable or linkable tables. Our core methods also generalize to other settings where computational tasks involve execution of programs or scripts.

In collaborative settings such as multi-investigator laboratories, data scientists need improved tools to manage not their data records but rather their data sets and data products, to facilitate both provenance tracking and data (and code) reuse within their data lakes and file systems. We demonstrate the Juneau System, which extends computational notebook software (Jupyter Notebook) as an instrumentation and data management point for overseeing and facilitating improved dataset usage, through capabilities for indexing, searching, and recommending "complementary" data sources, previously extracted machine learning features, and additional training data. This demonstration focuses on how we help the user find related datasets via search.

An increasing number of data science applications build on data lakes, which are schema-agnostic repositories of data files and data products. However data lakes typically offer limited capabilities for organizing and managing data. Data scientists and analysts need richer capabilities for querying for related tables and datasets in the lake, in order to find additional training data in similar tables, to find related and linkable tables, and to find other tables produced by the same workflows and thus similar in role. We focus on data science use cases within computational notebook software, such as Jupyter Notebook. However, our methods generalize to settings where computational tasks involve execution of programs or scripts. We explore how to find related tables, efficiently track a multitude of intermediate results and data products, and exploit similarity when storing related tables. PVLDB Reference Format: Yi Zhang and Zachary G. Ives. Finding Related Tables in the Data Lake. PVLDB, 13(xxx): xxxx-yyyy, 2019. DOI: https://doi.org/10.14778/xxxxxxx.xxxxxxx

The information revolution brought with it information pollution. Information retrieval and extraction help us cope with abundant information from diverse sources. But some sources are of anonymous authorship, and some are of uncertain accuracy, so how can we determine what we should actually believe? Not all information sources are equally trustworthy, and simply accepting the majority view is often wrong. This paper develops a general framework for estimating the trustworthiness of information sources in an environment where multiple sources provide claims and supporting evidence, and each claim can potentially be produced by multiple sources. We consider two settings: one in which information sources directly assert claims, and a more realistic and challenging one, in which claims are inferred from evidence provided by sources, via (possibly noisy) NLP techniques. Our key contribution is to develop a family of probabilistic models that jointly estimate the trustworthiness of sources, and the credibility of claims they assert. This is done while accounting for the (possibly noisy) NLP needed to infer claims from evidence supplied by sources. We evaluate our framework on several datasets, showing strong results and significant improvement over baselines.

Entity Suggestion with Conceptual Explanation (ESC) refers to a type of entity acquisition query in which a user provides a set of example entities as the query and obtains in return not only some related entities but also concepts which can best explain the query and the result. ESC is useful in many applications such as related-entity recommendation and query expansion. Many example based entity suggestion solutions are available in existing literatures. However, they are generally not aware of the concepts of query entities thus cannot be used for conceptual explanation. In this paper, we propose two probabilistic entity suggestion models and their computation solutions. Our models and solutions fully take advantage of the large scale taxonomies which consist of isA relations between entities and concepts. With our models and solutions, we can not only find the best entities to suggest but also derive the best concepts to explain the suggestion. Extensive evaluations on real data sets justify the accuracy of our models and the efficiency of our solutions.

Much work has focused on automatically constructing conceptual taxonomies or semantic networks from large text corpora. In this paper, we use a state-of-the-art data-driven conceptual taxonomy, Probase, to show that missing links in taxonomies are the chief problem that hinders their adoption by many real life applications, for the missing links break the inferencing that the conceptual taxonomy claims to support. To solve this problem, we devise a collaborative filtering framework to infer missing links in taxonomies derived from text corpora. We implement our method mainly on Probase, creating a denser taxonomy containing 5.1 million (about 30 percent) more isA relationships, with an accuracy of above 90 percent. We conduct comprehensive experiments to demonstrate the quality of the revised conceptual taxonomies.

Taxonomy is indispensable in understanding natural language. A variety of large scale, usage-based, data-driven lexical taxonomies have been constructed in recent years. Hypernym-hyponym relationship, which is considered as the backbone of lexical taxonomies can not only be used to categorize the data but also enables generalization. In particular, we focus on one of the most prominent properties of the hypernym-hyponym relationship, namely, transitivity, which has a significant implication for many applications. We show that, unlike human crafted ontologies and taxonomies, transitivity does not always hold in data-driven lexical taxonomies. We introduce a supervised approach to detect whether transitivity holds for any given pair of hypernym-hyponym relationships. Besides solving the inferencing problem, we also use the transitivity to derive new hypernym-hyponym relationships for data-driven lexical taxonomies. We conduct extensive experiments to show the effectiveness of our approach.

Knowledge base(KB) plays an important role in artificial intelligence. Much effort has been taken to both manually and automatically construct web-scale knowledge bases. Comparing with manually constructed KBs, automatically constructed KB is broader but with more noises. In this paper, we study the problem of improving the quality for automatically constructed web-scale knowledge bases, in particular, lexical taxonomies of isA relationships. We find that these taxonomies usually contain cycles, which are often introduced by incorrect isA relations. Inspired by this observation, we introduce two kinds of models to detect incorrect isA relations from cycles. The first one eliminates cycles by extracting directed acyclic graphs, and the other one eliminates cycles by grouping nodes into different levels. We implement our models on Probase, a state-of-the-art, automatically constructed, web-scale taxonomy. After processing tens of millions of relations, our models eliminate 74 thousand wrong relations with 91% accuracy.

Categories play a fundamental role in human cognition. Defining features (short for DFs) are the key elements to define a category, which enables machines to categorize objects. Categories enriched with their DFs significantly improve the machine's ability of categorization and benefit many applications built upon categorization. However, defining features can rarely be found for categories in current knowledge bases. Traditional efforts such as manual construction by domain experts are not practical to find defining features for millions of categories. In this paper, we make the first attempt to automatically find defining features for millions of categories in the real world. We formalize the defining feature learning problem and propose a bootstrapping solution to learn defining features from the features of entities belonging to a category. Experimental results show the effectiveness and efficiency of our method. Finally, we find defining features for overall 60,247 categories with acceptable accuracy.