In order to discover design principles for a large memory that can enable it to serve as the base of knowledge underlying human‐like language behavior, experiments with a model memory are being performed. This model is built up within a computer by “recoding” a body of information from an ordinary dictionary into a complex network of elements and associations interconnecting them. Then, the ability of a program to use the resulting model memory effectively for simulating human performance provides a test of its design. One simulation program, now running, is given the model memory and is required to compare and contrast the meanings of arbitrary pairs of English words. For each pair, the program locates any relevant semantic information within the model memory, draws inferences on the basis of this, and thereby discovers various relationships between the meanings of the two words. Finally, it creates English text to express its conclusions. The design principles embodied in the memory model, together with some of the methods used by the program, constitute a theory of how human memory for semantic and other conceptual material may be formatted, organized, and used.
The Teachable Language Comprehender (TLC) is a program designed to be capable of being taught to “comprehend” English text. When text which the program has not seen before is input to it, it comprehends that text by correctly relating each (explicit or implicit) assertion of the new text to a large memory. This memory is a “semantic network” representing factual assertions about the world.
The program also creates copies of the parts of its memory which have been found to relate to the new text, adapting and combining these copies to represent the meaning of the new text. By this means, the meaning of all text the program successfully comprehends is encoded into the same format as that of the memory. In this form it can be added into the memory.
Both factual assertions for the memory and the
capabilities
for correctly relating text to the memory's prior content are to be taught to the program as they are needed. TLC presently contains a relatively small number of examples of such assertions and capabilities, but within the system, notations for expressing either of these are provided. Thus the program now corresponds to a general process for comprehending language, and it provides a methodology for adding the additional information this process requires to actually comprehend text of any particular kind.
The memory structure and comprehension process of TLC allow new factual assertions and capabilities for relating text to such stored assertions to generalize automatically. That is, once such an assertion or capability is put into the system, it becomes available to help comprehend a great many other sentences in the future. Thus the addition of a single factual assertion or linguistic capability will often provide a large increment in TLC's effective knowledge of the world and in its overall ability to comprehend text.
The program's strategy is presented as a general theory of
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.