Computer assisted modeling of affective tone in written documents

Abstract

Beatie (1979) has expressed disappointment that the use of computers in literary scholarship since the nineteen-sixties has not had the positive impact that originally had been predicted. In fact, the main application of computer assistance has been to the rather straightforward tasks of preparing word counts and text concordances. Other procedures that have been developed to aid in literary research have suffered from the limitation of difficulty in application to studies other than the one for which the procedure was first developed. Thus the procedures have not become standard ones with which others could check the research findings themselves or readily apply to the examination of other materials.An attempt to identify general factors determining the attractiveness of aesthetic objects has been made in the work of Berlyne (1960, 1972). The identification of the motivational effects of variables like novelty, surprisingness, uncertainty, conflict, and complexity have permitted comparative studies. The resulting effects are examined in the context of arousal theory of motivation, according to which organisms strive to attain optimal levels of stimulation. In another paper, Dember (1965) argued that objects and events which have a powerful impact on the behavior of persons or rats are those with a complexity value somehow matched to the processing capabilities of the individual. He supports his argument with evidence relating to the choice by rats of striped walls over plain ones and the taste of persons for poems of intermediate complexity. Further studies of the relation between the attractiveness of pictures and designs and their objective complexity have been reported by Bertyne (1974).C.W. Anderson and G.E. McMaster are associate professors, respectively in the A complication in the extension of arousal level theories of stimulus attractiveness from pictures to stories is that the elements in literature are presented serially and are invariably rich in meaning. Measurement of characteristics of narrative that might define variables to be used for comparison, like novelty and surprisingness, must treat changes over time and be affected by the abstract qualities of the words, not their physical nature.One approach to the question of why some stories are more interesting and enduring than others has been outlined by Beaugrande and Colby (1979). They document the complexity of the old folktale "Tom Tit Tot" by diagramming the motivational states and actions of the main characters in a flowchart of possibilities from the initial problem condition of the story through to its conclusion. Emphasis is placed upon the amount of cognitive processing demanded by the rich texture of problems and motivations presented in the story. Depending heavily upon subjective judgment, this technique cannot be applied to voluminous bodies of literature.Another characteristic of engaging narrative is an ebb and flow of emotional tension, traditionally expected to build toward a gratifying release near the end of well-crafted storie...

Annotating Expressions of Opinions and Emotions in Language

“…Although many approaches to subjectivity classification focus only on the presence of subjectivity cue words themselves, disregarding context (e.g., Hart (1984), Anderson and McMaster (1982), Hatzivassiloglou and McKeown (1997), Turney (2002), Gordon et al (2003), Yi et al (2003)), the observations in this section suggest that different usages of words, in context, need to be distinguished to understand subjectivity.…”

“…Osgood et al (1957) and Heise (1965Heise ( , 2001, for example, aim to develop dimensional models of affect. In work by Hart (1984), Anderson and McMaster (1982), Biber and Finegan (1989), Subasic and Heuttner (2001), and others, the lexicons are used to automatically characterize political texts, literature, news, and other types of discourse, along various subjective lines. Recent work by Kaufer et al (2004) is noteworthy.…”

Does training increase the use of more emotionally laden words by nurses when talking with cancer patients? A randomised study

“…Counting emotional words may be a way to assess the emotional tone of a text (Osgood et al, 1957;Heise, 1965;Gottschalk and Bechtel, 1982). Emotional content may be also present in larger semantic unit such as utterances (Anderson and McMaster, 1982;Whissell and Dewson, 1986;Whissell et al, 1986;Martindale, 1987;Hogenraad and Bestgen, 1989;Bestgen, 1994). The comparison of the performance of content analysis systems and their respective benefit to assess the enhancement of HCP empathy by TW is still a matter of debate.…”

Modeling emotional tone in stories using tension levels and categorical states

“…Anderson andMcMaster (1982a, 1982b) have devised a method by which the emotional tone quality of text can be compared from unit to unit of a story and between the same sections of diff~erent stories. This system depends upon computer detection of those words which are among the most common 1000 in English (excluding the large number of function words found in any body of text).…”

“…The residual tension proposed by Anderson and McMaster (1982a) is given by the equation -E (evaluation) + A (activity). Considered in the light of the categorical state model, these scores reveal that high tension corresponds to one or another of two states, those that Mehrabian describes as anxious or hostile.…”