Semantically rich human-aided machine annotation

Abstract

This paper describes a semantically rich, human-aided machine annotation system created within the Ontological Semantics (OntoSem) environment using the DEKADE toolset. In contrast to mainstream annotation efforts, this method of annotation provides more information at a lower cost and, for the most part, shifts the maintenance of consistency to the system itself. In addition, each tagging effort not only produces knowledge resources for that corpus, but also leads to improvements in the knowledge environment that will better support subsequent tagging efforts.

Learning by Reading by Learning to Read

“…In parallel to work on unsupervised learning, we also plan to enhance our existing knowledge acquisition environment DEKADE [16], to include the option of presenting the results of automatic learning to human acquirers. This way we expect our work to contribute to the efficiency of human knowledge acquisition at an early stage.…”

An NLP Lexicon as a Largely Language-Independent Resource

“…On a day-to-day basis, this means that the unexpected input processor in our environment must work harder than it should have to, and that the production of "gold-standard" TMRs (which are, essentially, superannotated corpora; see McShane et al (2005a) for discussion) is slowed down by supplemental lexical acquisition. We already automatically generate and process unknown lexical items at run-time, using unidirectional selectional restrictions and accepting the resultant underspecification (see Nirenburg and Raskin 2004, pp. 279-292 for a sketch of this type of processing).…”

Introduction to Frontiers in Corpus Annotation II

“…Pie in the Sky can be viewed as a framework for unifying these annotation schemata into a single representation (a Unified Linguistic Annotation framework in the sense of (Pustejovsky et al, 2005)). Other work presented assumes that the integrated framework is the object of the annotation rather than the result of merging annotations (E. W. Hinrichs and S. Kübler and K. Naumann, 2005;McShane et al, 2005). There are pros and cons to both approaches.…”

“…(Poesio and Artstein, 2005) discusses some better ways of assessing interannotator agreement, particularly when there is a gray area between correct and incorrect annotation. (McShane et al, 2005) discusses the issue of human-aided annotation (human correction of a machine-generated analysis) as it pertains to a single-integrated annotation scheme, similar in many ways to "Pie in the Sky", although it has been in existence for a lot longer.…”

“…( McShane et al, 2005;Poesio and Artstein, 2005) (the ninth and tenth papers) both pertain to issues about improving the annotation process. (Poesio and Artstein, 2005) discusses some better ways of assessing interannotator agreement, particularly when there is a gray area between correct and incorrect annotation.…”