Research and Publications
Lymba's talented team of research scientists and natural language experts have collectively published more than 300 papers in the areas of semantics, ontologies, question answering, and reasoning. Our deep roots in research continue to drive the high quality of our products.
For more than 15 years, dating back to its time known as Language Computer Corporation, Lymba employees have participated in numerous government projects sponsored by ARDA/IARPA, Air Force, National Science Foundation, and Intelligence Community. State-of-the-art technologies resulting from these programs make it possible for Lymba products to provide best-in-class solutions.
Check out our selected publications:
This paper presents a novel approach to determine textual similarity. A layered methodology to transform text into logic forms is proposed, and semantic features are derived from a logic prover. Experimental results show that incorporating the semantic structure of sentences is beneficial. When training data is unavailable, scores obtained from the logic prover in an unsupervised manner outperform supervised methods.
Analysts in various domains, especially intelligence and financial, have to constantly extract useful knowledge from large amounts of unstructured or semi-structured data. Keyword-based search, faceted search, question-answering, etc. are some of the automated methodologies that have been used to help analysts in their tasks. General-purpose and domain-specific ontologies have been proposed to help these automated methods in organizing data and providing access to useful information. However, problems in ontology creation and maintenance have resulted in expensive procedures for expanding/maintaining the ontology library available to support the growing and evolving needs of analysts. In this paper, we present a generalized and improved procedure to automatically extract deep semantic information from text resources and rapidly create semantically-rich domain ontologies while keeping the manual intervention to a minimum. We also present evaluation results for the intelligence and financial ontology libraries, semi-automatically created by our proposed methodologies using freely-available textual resources from the Web.
The availability of massive amounts of raw domain data has created an urgent need for sophisticated AI systems with capabilities to find complex and useful information in big-data repositories in real-time. Such systems should have capabilities to process and extract significant information from natural language documents, search and answer complex questions, make sophisticated predictions about future events, and generally interact with users in much more powerful and intuitive ways. To be effective, these systems need a significant amount of domain-specific knowledge in addition to the general-domain knowledge. Ontologies/Knowledge-Bases represent knowledge about domains of interest and serve as the backbone for semantic technologies and applications. However, creating such domain models is time consuming, error prone, and the end product is difficult to maintain. In this paper, we present a novel methodology to automatically build semantically rich knowledge models for specific domains using domain-relevant unstructured data from resources such as web articles, manuals, e-books, blogs, etc. We also present evaluation results for our automatic ontology/knowledge-base generation methodology using freely-available textual resources from the World Wide Web.
Semantic representation of text is key to text understanding and reasoning. In this paper, we present Polaris, Lymba’s semantic parser. Polaris is a supervised semantic parser that given text extracts semantic relations. It extracts relations from a wide variety of lexico-syntactic patterns, including verb-argument structures, noun compounds and others. The output can be provided in several formats: XML, RDF triples, logic forms or plain text, facilitating interoperability with other tools. Polaris is implemented using eight separate modules. Each module is explained and a detailed example of processing using a sample sentence is provided. Overall results using a benchmark are discussed. Per module performance, including errors made and pruned by each module are also analyzed
This paper reports on Lymba Corporation’s (a spinoff of Language Computer Corporation) participation in the TREC 2007 Question Answering track. An overview of the PowerAnswer 4 question answering system and a discussion of new features added to meet the challenges of this year’s evaluation are detailed. Special attention was given to methods for incorporating blogs into the searchable collection, methods for improving answer precision, both statistical and knowledge driven, new mechanisms for recognizing named entities, events, and time expressions, and updated pattern-driven approaches to answer definition questions. Lymba’s results in the evaluation are presented at the end of the paper.
With the ever-growing adoption of AI technologies by large enterprises, purely data-driven approaches have dominated the field in the recent years. For a single use case, a development process looks simple: agreeing on an annotation schema, labeling the data, and training the models. As the number of use cases and their complexity increases, the development teams face issues with collective governance of the models, scalability and reusability of data and models. These issues are widely addressed on the engineering side, but not so much on the knowledge side. Ontologies have been a well-researched approach for capturing knowledge and can be used to augment a data-driven methodology. In this paper, we discuss 10 ways of leveraging ontologies for Natural Language Processing (NLP) and its applications. We use ontologies for rapid customization of a NLP pipeline, ontology related standards to power a rule engine and provide standard output format. We also discuss various use cases for medical, enterprise, financial, legal, and security domains, centered around three NLP-based applications: semantic search, question answering and natural language querying.
People convey sentiments and emotions through language. To understand these affectual states is an essential step towards understanding natural language. In this paper, we propose a transfer-learning based approach to inferring the affectual state of a person from their tweets. As opposed to the traditional machine learning models which require considerable effort in designing task specific features, our model can be well adapted to the proposed tasks with a very limited amount of fine-tuning, which significantly reduces the manual effort in feature engineering. We aim to show that by leveraging the pre-learned knowledge, transfer learning models can achieve competitive results in the affectual content analysis of tweets, compared to the traditional models. As shown by the experiments on SemEval-2018 Task 1: Affect in Tweets, our model ranking 2nd , 4th and 6th place in four of its subtasks proves the effectiveness of our idea.
Traditionally, the task of answering natural language questions has involved a keyword-based document retrieval step, followed by in-depth processing of candidate answer documents and paragraphs. This post-processing uses semantics to various degrees. In this article, we describe a purely semantic question answering (QA) framework for large document collections. Our high-precision approach transforms the semantic knowledge extracted from natural language texts into a language-agnostic RDF representation and indexes it into a scalable triplestore. In order to facilitate easy access to the information stored in the RDF semantic index, a user’s natural language questions are translated into SPARQL queries that return precise answers back to the user. The robustness of this framework is ensured by the natural language reasoning performed on the RDF store, by the query relaxation procedures, and the answer ranking techniques. The improvements in performance over a regular free text search index-based question answering engine prove that QA systems can benefit greatly from the addition and consumption of deep semantic information.