Automatic summarization extraction
This field is optional but you should use it to get better results
Enter your text:
Load a sample
#2744. 600 εκατ. ευρώ για τις καταστροφές
#2781. ΡΑΜΣΕΪ ΚΛΑΡΚ "Οι ΗΠΑ γεννούν την τρομοκρατία"
#2992. Όταν οι δικηγόροι… έγιναν δημοσιογράφοι
#3378. Mohamed Morsi's approach to Gaza air strikes falls short for many Egyptians
#4264. Jenson Button says 2012 McLaren car is worst since he joined F1 team
#5141. FBI probe of CIA chief David Petraeus's emails led to affair discovery – reports
TF: Term Frequency
ISF: Inverse Sentence Frequency, rewards the low frequency of words inside a document's sentences.
IDF: Inverse Document Frequency, rewards the low frequency of words inside a documents library.
RIDF: Residual IDF, rewards the low appearance probability of words inside our text.
TF-ISF is a lightweigh solution, TF-IDF & TF-RIDF use the internal documents and terms libraries to calculate the scores. TF-RIDF uses the Poison model to calculate the probabilities of the terms.
The sentences that exceed the words limits will be ingored. The 0 disables the limit.
The Article method ranks higher top paragraphs/sentence.
The Baxendale's method ranks higher the first and last sentences in a paragraph.
Sentence(i) Score = TW * T(i) + PW * P(i) + KW * K(i).
The T, P, K are each sentence's terms score, position score and keywords score.
The TW, PW, KW are the weights for linear combination of the three scores.
The weights are unsigned float values, set a weight to 0 to disable.
Page Generated in: 0.7789s, Memory Usage: 9.5774MB