Michael Glass

• PhD
• Researcher at IBM

We introduce a framework for generating Chain-of-Thought (CoT) rationales to enhance text-to-SQL model fine-tuning. These rationales consist of intermediate SQL statements and explanations, serving as incremental steps toward constructing the final SQL query. The process begins with manually annotating a small set of examples, which are then used t...

Text-to-SQL is emerging as a practical interface for real world databases. The dominant paradigm for Text-to-SQL is cross-database or schema-independent, supporting application schemas unseen during training. The schema of a database defines the tables, columns, column types and foreign key connections between tables. Real world schemas can be larg...

Increasing amounts of structured data can provide value for research and business if the relevant data can be located. Often the data is in a data lake without a consistent schema, making locating useful data challenging. Table search is a growing research area, but existing benchmarks have been limited to displayed tables. Tables sized and formatt...

Data preparation, also called data wrangling, is considered one of the most expensive and time-consuming steps when performing analytics or building machine learning models. Preparing data typically involves collecting and merging data from complex heterogeneous, and often large-scale data sources, such as data lakes. In this paper, we introduce a...

As demonstrated by GPT-3 and T5, transformers grow in capability as parameter spaces become larger and larger. However, for tasks that require a large amount of knowledge, non-parametric memory allows models to grow dramatically with a sub-linear increase in computational cost and GPU memory requirements. Recent models such as RAG and REALM have in...

In a recent work, we presented a novel state-of-the-art approach to zero-shot slot filling that extends dense passage retrieval with hard negatives and robust training procedures for retrieval augmented generation models. In this paper, we propose a system based on an enhanced version of this approach where we train task specific models for other k...

Most existing end-to-end Table Question Answering (Table QA) models consist of a two-stage framework with a retriever to select relevant table candidates from a corpus and a reader to locate the correct answers from table candidates. Even though the accuracy of the reader models is significantly improved with the recent transformer-based approaches...

In recent years, a number of keyphrase generation (KPG) approaches were proposed consisting of complex model architectures, dedicated training paradigms and decoding strategies. In this work, we opt for simplicity and show how a commonly used seq2seq language model, BART, can be easily adapted to generate keyphrases from the text in a single batch...

Automatically inducing high quality knowledge graphs from a given collection of documents still remains a challenging problem in AI. One way to make headway for this problem is through advancements in a related task known as slot filling. In this task, given an entity query in form of [Entity, Slot, ?], a system is asked to fill the slot by generat...

In this paper, we propose a fully automated system to extend knowledge graphs using external information from web-scale corpora. The designed system leverages a deep-learning-based technology for relation extraction that can be trained by a distantly supervised approach. In addition, the system uses a deep learning approach for knowledge base compl...

Recent advances in transformers have enabled Table Question Answering (Table QA) systems to achieve high accuracy and SOTA results on open domain datasets like WikiTableQuestions and WikiSQL. Such transformers are frequently pre-trained on open-domain content such as Wikipedia, where they effectively encode questions and corresponding tables from W...

We present the first end-to-end, transformer-based table question answering (QA) system that takes natural language questions and massive table corpus as inputs to retrieve the most relevant tables and locate the correct table cells to answer the question. Our system, CLTR, extends the current state-of-the-art QA over tables model to build an end-t...

The ability to automatically extract Knowledge Graphs (KG) from a given collection of documents is a long-standing problem in Artificial Intelligence. One way to assess this capability is through the task of slot filling. Given an entity query in form of [Entity, Slot, ?], a system is asked to `fill' the slot by generating or extracting the missing...

Transformer based architectures are recently used for the task of answering questions over tables. In order to improve the accuracy on this task, specialized pre-training techniques have been developed and applied on millions of open-domain web tables. In this paper, we propose two novel approaches demonstrating that one can achieve superior perfor...

Research on the task of Reading Comprehension style Question Answering (RCQA) has gained momentum in recent years due to the emergence of human annotated datasets and associated leaderboards, for example CoQA, HotpotQA, SQuAD, TriviaQA, etc. While state-of-the-art has advanced considerably, there is still ample opportunity to advance it further on...

We introduce Grinch, a new algorithm for large-scale, non-greedy hierarchical clustering with general linkage functions that compute arbitrary similarity between two point sets. The key components of Grinch are its rotate and graft subroutines that efficiently reconfigure the hierarchy as new points arrive, supporting discovery of clusters with com...

Existing literature on Question Answering (QA) mostly focuses on algorithmic novelty, data augmentation, or increasingly large pre-trained language models like XLNet and RoBERTa. Additionally, a lot of systems on the QA leaderboards do not have associated research documentation in order to successfully replicate their experiments. In this paper, we...

BERT (Bidirectional Encoder Representations from Transformers) and related pre-trained Transformers have provided large gains across many language understanding tasks, achieving a new state-of-the-art (SOTA). BERT is pre-trained on two auxiliary tasks: Masked Language Model and Next Sentence Prediction. In this paper we introduce a new pre-training...

In this paper, we propose a fully automated system to extend knowledge graphs using external information from web-scale corpora. The designed system leverages a deep learning based technology for relation extraction that can be trained by a distantly supervised approach. In addition to that, the system uses a deep learning approach for knowledge ba...

Transfer learning enhances learning across tasks, by leveraging previously learned representations -- if they are properly chosen. We describe an efficient method to accurately estimate the appropriateness of a previously trained model for use in a new learning task. We use this measure, which we call "Predict To Learn" ("P2L"), in the two very dif...

This paper introduces a novel orchestration framework, called CFO (COMPUTATION FLOW ORCHESTRATOR), for building, experimenting with, and deploying interactive NLP (Natural Language Processing) and IR (Information Retrieval) systems to production environments. We then demonstrate a question answering system built using this framework which incorpora...

We introduce Grinch, a new algorithm for large-scale, non-greedy hierarchical clustering with general linkage functions that compute arbitrary similarity between two point sets. The key components of Grinch are its rotate and graft subroutines that efficiently reconfigure the hierarchy as new points arrive, supporting discovery of clusters with com...

We propose a novel approach to learn representations of relations expressed by their textual mentions. In our assumption, if two pairs of entities belong to the same relation, then those two pairs are analogous. We collect a large set of analogous pairs by matching triples in knowledge bases with web-scale corpora through distant supervision. This...

We address relation extraction as an analogy problem by proposing a novel approach to learn representations of relations expressed by their textual mentions. In our assumption, if two pairs of entities belong to the same relation, then those two pairs are analogous. Following this idea, we collect a large set of analogous pairs by matching triples...

We propose a novel approach to learn representations of relations expressed by their textual mentions. In our assumption, if two pairs of entities belonging to the same relation, then those two pairs are analogous. We collect a large set of analogous pairs by matching triples in knowledge bases with web-scale corpora through distant supervision. Th...

Knowledge Base Population (KBP) is an important problem in Semantic Web research and a key requirement for successful adoption of semantic technologies in many applications. In this paper we present Socrates, a deep learning based solution for Automated Knowledge Base Population from Text. Socrates does not require manual annotations which would ma...

For many domains, structured knowledge is in short supply, while unstructured text is plentiful. Knowledge Base Population (KBP) is the task of building or extending a knowledge base from text, and systems for KBP have grown in capability and scope. However, existing datasets for KBP are all limited by multiple issues: small in size, not open or ac...

This paper addresses automatic extraction of abbreviations (encompassing acronyms and initialisms) and corresponding long-form expansions from plain unstructured text. We create and are going to release a multilingual resource for abbreviations and their corresponding expansions, built automatically by exploiting Wikipedia redirect and disambiguati...

We present WatsonPaths, a novel system that can answer scenario-based questions. These include medical questions that present a patient summary and ask for the most likely diagnosis or most appropriate treatment. Watson-Paths builds on the IBM Watson question-answering system. WatsonPaths breaks down the input scenario into individual pieces of inf...

We apply a general deep learning framework to address the non-factoid question answering task. Our approach does not rely on any linguistic tools and can be applied to different languages or domains. Various architectures are presented and compared. We create and release a QA corpus and setup a new QA task in the insurance domain. Experimental resu...

In this work we present a novel approach to bootstrap domain specific terminology, namely Structured Term Recognition, and we apply it to the medical domain. In contrast to previous approaches, based on observing distributional properties of terminology with respect to their contexts, our method analyzes the "internal structure" of multi-word terms...

Relation extraction is the task of finding entities in text connected by semantic relations. Bootstrap-ping approaches to relation extraction have gained considerable attention in recent years. These ap-proaches are built with an underlying assumption, that when a pair of words is known to be related in a specific way, sentences containing those wo...

Markov logic networks have been successfully applied to many problems in AI. However, the computational complexity of the inference procedures has limited their application. Previous work in lifted inference, lazy inference and cutting plane inference has identified cases where the entire ground network need not be constructed. These approaches are...

Extracting a formal representation from text that can be used to reason and answer questions has long been a goal of Artifi-cial Intelligence research. We demonstrate a method for knowledge engineers to con-struct a semantic interpeter that requires little natural language processing exper-tise. The resulting semantic interpreter is also able to ex...

When extending a scientific knowledge base with new information, particularly information presented in nat- ural language, it is important that the information be encoded in a form that is compatible with the existing knowledge base. Hand built systems for semantic in- terpretation and knowledge integration can suffer from brittleness. Methods for...

A traditional goal of Artificial Intelligence resea rch has been a system that can read unrestricted natural la nguage texts on a given topic, build a model of that topic and reason over the model. Natural Language Processing advances in syntax and semantics have made it possible to extra ct a limited form of meaning from sentences. Knowledge Repre...