THE INTEGRATION OF BANKING AND TELECOMMUNICATIONS: THE NEED FOR REGULATORY REFORM

(Jeff_L) #1
AUTOMATED AUTHORSHIP ANALYSIS 307

Figure 4. System diagram for Modality. Note the four parallel taxonomies.


II. EXPERIMENTS IN AUTHORSHIP PROFILING

The uses of these features can be seen both in authorship
attribution and in authorship profiling, where we seek to
determine characteristics of a text’s author (such as sex, age, or
personality), even in the absence of any specific candidate
authors. We describe here some experiments we have done on
authorship profiling for author sex, age, native language, and
personality.^12
In these experiments, we compared the use of functional
lexical features as above with content-based features, namely,


(^12) See Shlomo Argamon et al., Automatically Profiling the Author of an
Anonymous Text, COMM. ACM, Feb. 2009, at 119.

Free download pdf