/*
ComputeImpact return percentage of words in new revision compared to old
revision, using
count_of_words_in_new
/
(count_of_words_in_old + count_of_words_in_new)
if no words are found in old and new revision, return 0.
*/
func ComputeImpact(oldrevid, newrevid string, wordlist []string) float64 {
oldtext, _ := revision.GetContentClean(oldrevid)
newtext, _ := revision.GetContentClean(newrevid)
oldCnt := tekstus.StringCountTokens(oldtext, wordlist, false)
newCnt := tekstus.StringCountTokens(newtext, wordlist, false)
total := float64(oldCnt + newCnt)
if total == 0 {
return 0
}
return float64(newCnt) / total
}
/*
Compute the frequency of inserted words.
*/
func (ftr *TermFrequency) Compute(dataset tabula.DatasetInterface) {
newrevidx := dataset.GetColumnByName("newrevisionid")
adds := dataset.GetColumnByName("additions")
recordslen := len(adds.Records)
for x, rec := range adds.Records {
r := tabula.NewRecordReal(float64(0))
// Get inserted words.
intext := rec.String()
if len(intext) == 0 {
ftr.PushBack(r)
continue
}
intext = clean.WikiText(intext)
inWords := tekstus.StringSplitWords(intext, true, true)
// Get content of new revision.
revid := newrevidx.Records[x].String()
if DEBUG >= 2 {
fmt.Printf("[feature] term_frequency: %d/%d processing %q\n",
x, recordslen, revid)
}
newtext, e := revision.GetContentClean(revid)
if e != nil {
ftr.PushBack(r)
continue
}
newWords := tekstus.StringSplitWords(newtext, true, false)
freq := tekstus.WordsFrequenciesOf(newWords, inWords, false)
r.SetFloat(Round(freq))
ftr.PushBack(r)
}
}
