Past Projects

Hover over a project to read a quick overview; click on the project to read a complete description.

• Multiclass Classification
• Context Sensitive Paraphrasing
• Semantic Role Labeling with Psycholinguistic Inspired Features
• Semisupervised Learning with Prior Knowledge
• Focused Textual Entailment
• Digital Guards for Open System Security
• Structural Decomposition of Sentences
• Constraint Driven Learning
• Robust Adaptation: Analysis of Feature Extension Algorithms
• Relation Extraction with Auxilliary Tasks
• Bootstrap Learning
• PhraseNet

  The purpose of PhraseNet is to build a context-sensitive lexical semantic knowledge system, which can help various Natural Language Processing tasks such as question answering and prepositional phrase attachment.

• Digital Guards for Open System Security
• Named Entity Recognition

  Understanding text to the level that we can extract information from it in an intelligent way and answer questions with respect to it requires the ability to identify different types of entities and categories in text. E.g., this phrase represents a name

• Name Identification and Tracing

  A given entity - representing a person, a location, or an organization - may be mentioned in text in multiple, ambiguous ways. Understanding natural language and supporting intelligent access to textual information require identifying whether different mentions of a name, within and across documents, represents the same entity

• Question Answering Systems

  As the amount of information grows on the web, it becomes harder to find information. Research on Question Answering Systems aims at making the task of finding information easier. The goal is to replace current search technologies, which are based solely

• Intermediate Knowledge Representations that Facilitate Learning
• Inference with Classifiers
• Learning Coherent Concepts
• Constraint Classifications: A New Approach to Multiclass Classification
• Learning Sparse Representations for Object Detection
• Kindle: Knowledge and INference via Description Logics for natural languagE
• Reflex: Named Entity Recognition and Transliteration for 50 Languages
• Unsupervised Rank Aggregation

  The need to meaningfully combine sets of rankings often comes up when one deals with ranked data. Although a number of heuristic and supervised learning approaches to rank aggregation exist, they require domain knowledge or supervised ranked data, both of which are expensive to acquire. We investigate learning methods for aggregation of (partial) rankings without supervision.