Evaluating and comparing three text-production techniques

Abstract

What al+c the benefits of ttsing Natural [~anguagc (;cneratio,t in an industrial apl+lication? We have attempt t<) answer part (}f this qttcsti{}n with at descripti(}n of an assessment {}f three techniques for producing multiscntcntial text: sentiatutomatic fill-in-lhc-blank interfacing, automalic linguistic-and-tcmphltes hybrid generation, and hunlall writing. This asscssIllol]l used a black b(}x motlmdology, with ain independetlt blindtested jury that gave difforent quality levels in relation to a sot o1' criteria. The texts used for tile assessliicnt wcfc business reply letters.

Natural language processing in mental health applications using non-clinical texts

“…Evaluations have shown that this type of semi-automatic and fully-automatic texts (i.e. letter, interventions) compare well with those generated by humans on a number of measures such as tone, rhythm and flow, repetition and terminology (Coch 1996). Within health, personalization has been considered critical to patient-centered care and a number of studies have evaluated Natural language Generation (NLG) techniques in the authoring and personalization of webpages containing patient education materials.…”

On the role of linguistic descriptions of data in the building of natural language generation systems

“…Another example of how NLG solutions can lighten tedious tasks is the AlethGen engine [16,50], which was used to automatically generate letters to answer customer issues and questions for the French largest mail order company, La Redoute. The developed system used data introduced by a human operator, a customer database and knowledge bases to generate an answering text for a customer request.…”

Building applied natural language generation systems

“…Examples of using nlg as an authoring aid include: helping customer service representatives write letters for customers, as in (Springer et al 1991Coch 1996 helping engineers produce management summaries of design paths they have explored, as in (McKeown et al 1994) helping personnel o cers produce job descriptions, as in (Caldwell and Korelsky 1994) and helping technical authors produce instructions for using software, as in (Paris et al 1995). Again, this list is indicative o n l y .…”