Gravity wells of meaning: detecting information‐rich passages in scientific texts

Abstract

Four term‐weighting schemes are used to detect information‐rich passages in texts and the results are compared. It is demonstrated that word categories and frequency‐derived weights have a close correlation but that weighting according to the first mention theory or the cue‐method shows no correlation with frequency‐based weights.

Citation

Paijmans, H. (1997), "Gravity wells of meaning: detecting information‐rich passages in scientific texts", Journal of Documentation, Vol. 53 No. 5, pp. 520-536. https://doi.org/10.1108/EUM0000000007209