LREC 2014 Tutorial Deep Semantic Annotation with Shallow Methods Presenters: James Pustejovsky and Anna Rumshisky 9:00-9:30am: Part I. Introduction to Deep Semantic Annotation (DSA). 1. Differences between Shallow and Deep Annotation: what makes a semantic annotation "deep"? a. No overt syntactic presence b. No overt semantic context 9:30-10:30am: Part II. Kinds of Deep Semantic Annotation. 1. Covert Entity Recognition a. Entailed implicit arguments b. Presupposed arguments, properties, or propositions c. Recovering Coercion Contexts 2. Covert Relation Recognition a. Semantic Roles b. Temporal Relations c. Intersentential Discourse Relations 10:30-11:00am: Coffee Break 11:00-12:30pm: Part III. Task Decomposition for Specific Deep Tasks. 1. Introduction to Task Decomposition with Mechanical Turk 2. Detailed Example Task Decompositions a. Argument Selection and Type Coercion (including Qualia modification) b. Temporal Relation identification (TIMEX3 recognition and event ordering and anchoring) e. Graded Sense Clustering (using pairwise similarity)