AI Approach For Automatic Text Summarization

Abstract

æ Ag Ég ñëð , ú

Í

B@ ñ

JË@ J

jÊ

K ÈAm . ×

ékðQ£

B@ è

Yë ÈðA

J

K . A

J

ÓñK

Aë

ðA

@

Õ

æK

ú

æË@

éJ

JË@

HA

KAJ

J . Ë@

áÓ

éÊ

KAêË@

éJ

ÒºË@

èP@X@

ø

Yj

, JË

éJ

k . @Q

j

J B@ I . J

ËA

B@ ½Ë

X ú

¯ AÖ ß . ,

é

®Ê

J

jÖ Ï @ J

jÊ

JË@

HAJ

J

®

K

­

º

J

AêÒJ

J

®

Kð

éJ

ËAm Ì '@ h .

XAÒ

JÊË

éÊÓA

éªk . @QÓ A

J

@PX ÉÒ

.

é

JJ

j . êË@ð ,

éK

YK

Qj .

JË@

á

j

JË

HAJ

J

®

JË@ è

Yë ú

¯

èñ

®Ë@ A

®

K

á

K . ©Òm . '

A

J

m × @

PA£@

h Q

®

K .

éJ

ËAª

¯

HA

KAJ

K .

é«ñÒm . × úÎ «

IK

Qk .

@ ú

æË@ H . PAj .

JË@ Qê

¢

. J

jÊ

JË@

èXñk . ú

¯

á

®Ë@ øñ

J Ö ß . ©

¯YK

AÜ Ø ,

éº AÒ

JÓð

è

Qk . ñÓ

HA

jÊÓ YJ

Ëñ

K ú

This thesis investigates the field of automatic text summarization, a crucial solution to the challenge of managing the vast amount of textual data generated daily. We explore various summarization techniques, including extractive, abstractive, and hybrid approaches. Our study includes a thorough review of existing models and their evaluation. We propose an enhanced framework that combines the strengths of these techniques to improve summarization quality. Experiments conducted on the CNN/DailyMail dataset demonstrate the effectiveness of our approach in generating concise and coherent summaries, advancing the state-of-the-art in text summarization

¯ A

Jj . î

E .