Pausanias Analysis

Tri-marked sentence-level mythic vs. historical analysis

Surface, including books 4 and 8, with rhetoric markers

This model drops sentences tagged other and fits a balanced TF-IDF logistic regression to the remaining mythic and historical tags.

Model Performance Metrics

The following metrics are from the logistic regression classifier's performance on the test set:

Class Precision Recall F1-Score Support
Historical 0.703 0.750 0.726 859
Mythic 0.745 0.697 0.720 899
Overall Accuracy 0.723 1758

Confusion Matrix

Predicted
Historical Mythic
Actual Historical 644 215
Mythic 272 627

Counts

7,032mythic/historical sentences
3,598mythic
3,434historical
1,000features

Predictors of Mythic Sentences

Sort by:
Word/Phrase English Coefficient Mythic Count Historical Count p-value q-value
λέγουσιν they say 2.8403 223 47 4.1e-28 2.05e-25
φασὶν they say 2.6694 111 20 3.33e-16 6.67e-14
εἶναι to be 2.6537 430 138 7.46e-36 7.46e-33
φασιν they say 2.6434 170 31 5.38e-24 1.79e-21
ἔπη words 2.1104 49 2 1.67e-12 1.39e-10
ἀφικέσθαι to arrive 2.0922 42 4 6.51e-09 3.25e-07
ὄνομα name 2.0818 181 54 1.08e-16 2.7e-14
ὕδωρ water 1.9995 57 11 1.35e-08 5.02e-07
αὐτὴν her 1.8972 110 39 1.09e-08 4.19e-07
λέγουσι they say 1.8252 131 31 1.89e-15 2.7e-13
ἐστιν is 1.7894 258 116 7.02e-13 6.38e-11
αὐτήν her 1.7126 33 5 2.86e-06 5.97e-05
θυγατέρα daughter 1.7067 82 9 8.31e-16 1.38e-13
γυναῖκα woman; wife 1.6995 59 14 1.19e-07 3.2e-06
ἀνθρώπων of humans 1.6636 54 25 0.00185 0.0092
φασι they say 1.6601 51 8 8.17e-09 3.55e-07
ἐστὶν is 1.6401 140 65 4.61e-07 1.15e-05
λέγοντεσ speaking 1.6280 31 3 8.81e-07 2.1e-05
καλεῖται is called 1.5756 19 1 4.25e-05 0.000489
καὶ εἶναι and to be 1.5429 37 1 3.54e-10 2.21e-08
ἔστι there is 1.5364 99 36 1.08e-07 3.16e-06
βωμὸσ altar 1.5203 19 0 4.08e-06 7.84e-05
δʼ but; and 1.4846 109 34 4.95e-10 2.91e-08
θηρίον wild animal 1.4844 17 0 1.61e-05 0.000215
πεποίηται has been made 1.4149 57 23 0.000235 0.00197
αὕτη this (feminine) 1.3997 19 9 0.0731 0.159
ἐστι is 1.3822 110 41 3.86e-08 1.29e-06
θυγατρὶ to/for the daughter 1.3320 18 0 8.1e-06 0.000137
λόγοσ word; speech; account; reason; argument 1.3295 56 16 2.84e-06 5.97e-05
λόγον word 1.3286 41 17 0.00239 0.0116

Predictors of Historical Sentences

Sort by:
Word/Phrase English Coefficient Mythic Count Historical Count p-value q-value
ἐναντία opposite; contrary -2.1056 2 48 4.44e-13 4.44e-11
πόλεμον war -2.0280 12 80 8.47e-15 1.06e-12
πολέμωι in war; by/with war -2.0239 7 53 7.25e-11 5.18e-09
αὖθισ again -1.9846 35 100 1.6e-09 8.4e-08
εἰκόνα image -1.8659 5 33 7.31e-07 1.78e-05
χρημάτων of money; of goods -1.8068 5 36 1.18e-07 3.2e-06
ἐπίγραμμα epigram; inscription -1.7714 15 34 0.00348 0.0152
χρήματα money -1.7287 4 39 9.42e-09 3.92e-07
ἔργου of work -1.7246 4 35 1.11e-07 3.16e-06
πολέμου of war -1.6825 21 73 7.7e-09 3.55e-07
στρατιᾶι to/for the army -1.6524 3 27 3.04e-06 6.14e-05
τούσ the -1.5985 4 28 7.44e-06 0.00013
ἐν μάχηι in battle -1.5180 10 24 0.00996 0.0342
σφῶν of them; their -1.5180 2 16 0.000565 0.00368
ἀναστάτουσ expelled -1.5138 1 21 3.7e-06 7.25e-05
ὅσον as much as -1.5103 17 43 0.000311 0.00243
χαλκοῦν of bronze; brazen -1.5015 8 24 0.00245 0.0117
ἄνδρασ men -1.4851 16 53 1.71e-06 3.88e-05
βασιλέων of kings -1.4605 3 19 0.000375 0.00278
συμμαχίαν alliance -1.4539 1 19 1.42e-05 0.000203
τούτουσ these -1.4533 14 43 3.77e-05 0.000449
ἑκατόν one hundred -1.4476 0 16 1.03e-05 0.000166
βουλῆσ of the council; of the will/plan -1.4381 1 18 2.78e-05 0.000348
πυγμῆι with the fist -1.4256 1 14 0.000396 0.00281
καὶ περὶ and about -1.3804 9 26 0.00208 0.0103
ἐνίκα he/she/it was winning -1.3714 1 30 7.95e-09 3.55e-07
βαρβάρων of the barbarians -1.3708 7 32 1.58e-05 0.000215
ἐγένετο it happened -1.3609 61 118 3.07e-06 6.14e-05
πλέον more -1.3604 18 51 1.99e-05 0.000259
ὕστερον later; afterwards -1.3436 113 200 4.2e-08 1.35e-06