Marjorie Mcshane - Linguistics for the Age of AI

High-level sketch of the OntoAgent architecture.

Horizontal and vertical incrementality.

Stages of vertical context available during NLU by LEIAs.

The control flow of decision-making during semantic analysis.

Decision points during vertical-incremental processing.

The constituency parse for A fox caught a rabbit.

The dependency parse for A fox caught a rabbit.

A visual representation of syn-mapping. For the input He ate a sandwich, eat-v1 is a good match because all syntactic expectations are satisfied by elements of input. Eat-v2 is not a good match because the required words away and at are not attested in the input.

The processing flow involving syn-mapping. If the initial parse generates at least one perfect syn-map, then the agent proceeds along the normal course of analysis (stages 36: Basic Semantic Analysis, Basic Coreference Resolution, Extended Semantic Analysis, and Situational Reasoning). If it does not, then two recovery strategies are attempted, followed by reparsing. If the new parse is perfect, then the agent proceeds normally (stages 36). By contrast, if the new parse is also imperfect, the agent decides whether to optimize the available syn-maps and proceed normally (stages 36) or skip stages 35 and jump directly to stage 6, Situational Reasoning, where computing semantics with minimal syntax will be attempted.

A subset of paired, syntactically identical senses.

A more detailed view of the OntoAgent architecture than the one presented in .

The Maryland Virtual Patient (MVP) architecture.

This is a subset of the binding sets that use eat-v1 to analyze the input Cakeno, chocolate cakeId eat every day. The ellipses in the last row indicate that many more binding sets are actually generated, including even a set that leaves everything unbound, since this computational approach involves generating every possibility and then discarding all but the highest-scoring ones.

Types of modality used in Ontological Semantics

Examples of syntactic components of tag-question constructions

Comparison of best-case analyses of NNs across paradigms

.

VP ellipsis constructions

Referential and nonreferential uses of the same types of categories

Ellipsis-resolved meaning representation for John washed his car yesterday but Jane didnt

Canonical and variable-inclusive forms of idioms recorded as different lexical senses

Classes of comparative examples and when they are treated during NLU

Sample GERD levels and associated properties

Computing, rather than asserting, why patients have different end stages of GERD. Column 2 indicates each patients MODIFIED-TOTAL-TIME-IN-ACID- REFLUX per day. The cells in the remaining columns indicate the total time in acid reflux needed for GERD to advance in that stage. Cells with gray shading indicate that the disease will not advance to this stage unless the patients MODIFIED-TOTAL- TIME-IN-ACID-REFLUX changeswhich could occur, for example, if the patient took certain types of medications, changed its lifestyle habits, or had certain kinds of surgery.

Modeling complete and partial responses to medications. The reduction in MODIFIED-TOTAL- TIME-IN-ACID-REFLUX is listed first, followed by the resulting MODIFIED-TOTAL-TIME-IN-ACID- REFLUX in brackets.

Learning lexicon and ontology through language interaction

Patient-authoring choices for the disease achalasia

Examples of ontological knowledge about tests relevant for achalasia

Examples of knowledge that supports clinical decision-making about achalasia, which is used by the virtual tutor in the MVP system

MVP. Knowledge about the test results expected at different stages of the disease achalasia. Used by the tutoring agent in MVP. The test results in italics are required to definitively diagnose the disease.

Inventory of under-the-hood panes that are dynamically populated during MVP simulation runs

Examples of properties, associated with their respective concepts, whose values can potentially be automatically learned from the literature

Fast-lane elicitation strategy for recording information about physiology and symptoms

Sample precondition of good practice. Domain experts supply the descriptive fillers and knowledge engineers convert it into a formal representation.

Functionalities of a bias-detection advisor in clinical medicine

Four clinical properties of the esophageal disease achalasia, with values written in plain English for readability

Knowledge about expected test results during progression of achalasia

Example of halo-property nests

Examples of constructions that can lead to biased thinking

Learning while assembling the right back leg