TASC has recently developed a technology for learning scenario-templateextraction grammars from examples provided by end-users with no special computational or linguistic knowledge. For straightforward scenario-template problems, complete extraction systems can be learned from scratch in a few hours. The learned systems are fast, robust, and as accurate as carefully handcrafted systems. For mor...

The CoNLL-2012 shared task is an extension of the last year’s coreference task. We participated in the closed track of the shared tasks in both years. In this paper, we present the improvements of Illinois-Coref system from last year. We focus on improving mention detection and pronoun coreference resolution, and present a new learning protocol. These new strategies boost the performance of the...

Scoring the performance of a system is an extremely important aspect of coreference algorithm performance. The score for a particular run is the single strongest measure of how well the system is performing and it can strongly determine directions for further improvements. In this paper, we present several di erent scoring algorithms and detail their respective strengths and weaknesses for vary...

We present an annotation schema as part of an effort to create a manually annotated corpus for Arabic dialogue language understanding including spoken dialogue and written „chat‟ dialogue for inquiryanswer domain. The proposed schema handles mainly the request and response acts that occurs frequently in inquiry-answer debate conversations expressing request services, suggests, and offers. We ap...

This paper examines the use of co-switching in Navajo written poetry. I look specifically at code-switching from English dominant poems to Navajo. outline three general semantic domains that are most commonly code-switched Navajo: 1) emotions; 2) mythic characters; and 3) traditional place-names. suggest this has do with a linguistic ideology understands these as incommensurate English. argue s...

Statistical Language Understanding Using Frame Semantics

This is a very young project, operational for only a few months . The focus of the effort is dataextraction, the identification of instances of data-classes in "commercial" text -e .g., newspapers, technical reports, business correspondence, intelligence briefs . Instances of data-classes in text are phrases which identify factual content, such as names of people or organizations, products, fin...