ROLE: CONTEXTUAL VOCABULARY ACQUISITION: From Algorithm to Curriculum

1. Since 7/2001, funded by an NSF ROLE pilot project, William J. Rapaport (Computer Science & Engineering) and Michael W. Kibby (Learning & Instruction) have been investigating “contextual vocabulary acquisition” (CVA): active, deliberate acquisition of word meanings from text by reasoning from contextual clues, prior knowledge, language knowledge, and hypotheses developed from prior encounters with the word, but without external sources of help such as dictionaries or people. The intellectual merit of the project lies in increasing our understanding (based on case studies—“think-aloud” protocols) of how good readers use CVA to hypothesize a sense for an unknown word encountered in written context, using these observations to extend the computational theory of CVA, developing further a computer program that implements and tests this theory, and creating and evaluating a curriculum (based on the computational theory) to improve students’ abilities to use CVA. The broader impact is not merely to improve vocabulary acquisition, but also to increase reading comprehension of STEM texts, thereby leading to increased learning, and, by using a “miniature” (but real) example of the scientific method (viz., CVA), to give students a better understanding of the scientific method. The computational and educational strands of the research are fully integrated and jointly serve these goals. The research falls within Quadrants 2 and 3. 2. People know the meanings of more words than they are explicitly taught, so they must have learned most of them as a by-product of reading or listening. Some of this is the result of active processes of hypothesizing meanings for unknown words from context. How do readers do this? Most published strategies are quite vague; one simply suggests to “look” and “guess”. This vagueness stems from a lack of relevant research on the reasoning and language processes used for CVA. There is no generally accepted cognitive theory of CVA, nor is there an educational curriculum or set of detailed strategies for teaching it. If we knew more about how context operates, had a better theory of CVA, and knew how to teach it, we could more effectively help students identify context cues and know better how to use them. Artificial-intelligence and computational-linguistic studies of CVA (including ours) have necessarily gone into much more detail on what underlies the unhelpful advice to “guess”, since natural-language-processing (NLP) systems must operate on unconstrained input text independently of humans and can’t assume a “fixed complete lexicon”. But such studies have largely been designed to improve practical NLP systems. Few, if any, have been applied in an educational setting, and virtually all have been ignored in the readingand vocabulary-education literature. AI algorithms for CVA can fill in the details that can turn “guessing” into “computing”; these can then be taught to students. This multidisciplinary proposal combines basic and applied research to address twin needs: for NLP systems that operate independently of human assistance and for improving students’ abilities to read STEM materials. Its theoretical significance comes from the development of an NLP system that does CVA. Its educational significance lies in whether the knowledge gained from this system can be applied to teaching CVA strategies so that students can use them successfully when they encounter hard words in their regular (STEM) reading. The mutual feedback between the development of the computational theory based on the education team’s data and of the algorithm-based educational curriculum is also distinctive, making this a true cognitive-science project. 3. The AI team will: continue developing noun, verb, and adjective algorithms using insights from the thinkaloud protocols produced by the education team, develop an NL explanation facility for the system, develop an NL input/output system, investigate the use of “internal” context (morphology) for developing definitions, investigate the possibility of using OpenCYC as one source of general background information, and improve the current definitionrevision system. The education team will design and evaluate a CVA curriculum for middleand secondary-school students designed to increase: reading comprehension, word consciousness, CVA strategies based on the algorithms, meaning vocabulary, assessments of sense of meaning of hard words, and interest in words. The curriculum will teach students to: focus attention on a hard word in a text (word consciousness) and the textual clues to the word’s meaning; connect these clues with their language knowledge, reasoning processes, and prior knowledge to hypothesize a contextual