Problem klasifikacije programskog koda ; Source Code Classification Problem
Gespeichert in:
| Titel: | Problem klasifikacije programskog koda ; Source Code Classification Problem |
|---|---|
| Autoren: | Patajac, Matija |
| Weitere Verfasser: | Matijević, Domagoj |
| Verlagsinformationen: | Sveučilište Josipa Jurja Strossmayera u Osijeku. Odjel za matematiku. Josip Juraj Strossmayer University of Osijek. Department of Mathematics. |
| Publikationsjahr: | 2021 |
| Bestand: | Croatian Digital Theses Repository (National and University Library in Zagreb) |
| Schlagwörter: | PRIRODNE ZNANOSTI. Matematika. Matematička logika i računarstvo, NATURAL SCIENCES. Mathematics. Mathematical Logic and Accounting, programski kod, obrada programskog jezika, neuronske mreže, LSTM, vektorizacija, attention mehanizam, POJ-104, Source code, programming language processing, neural networks, embedding, attention mechanism |
| Beschreibung: | U ovom radu bavit ćemo se problemom klasifikacije programskog koda, gdje nam je cilj pomoću obrade programskog jezika odrediti funkcionalnost nekog isječka koda. Predložit ćemo rješenje u obliku neuronske mreže bazirane na long short-term memory (LSTM) rekurentnoj mreži te dobivene rezultate usporediti s nekoliko radova koji su se bavili ovom temom. ; In this paper we discuss source code classification problem, the goal of which is to determine the functionality of a code snippet using programming language processing. We propose a solution in the form of a neural network based on the long short-term memory (LSTM) recurrent network and compare our results to several other papers that had discussed this topic. |
| Publikationsart: | bachelor thesis |
| Dateibeschreibung: | application/pdf |
| Sprache: | Croatian |
| Relation: | https://zir.nsk.hr/islandora/object/mathos:598; https://urn.nsk.hr/urn:nbn:hr:126:228108; https://repozitorij.unios.hr/islandora/object/mathos:598; https://repozitorij.unios.hr/islandora/object/mathos:598/datastream/PDF |
| Verfügbarkeit: | https://zir.nsk.hr/islandora/object/mathos:598 https://urn.nsk.hr/urn:nbn:hr:126:228108 https://repozitorij.unios.hr/islandora/object/mathos:598 https://repozitorij.unios.hr/islandora/object/mathos:598/datastream/PDF |
| Rights: | http://rightsstatements.org/vocab/InC/1.0/ ; info:eu-repo/semantics/openAccess |
| Dokumentencode: | edsbas.4870243F |
| Datenbank: | BASE |
Nájsť tento článok vo Web of Science | Abstract: | U ovom radu bavit ćemo se problemom klasifikacije programskog koda, gdje nam je cilj pomoću obrade programskog jezika odrediti funkcionalnost nekog isječka koda. Predložit ćemo rješenje u obliku neuronske mreže bazirane na long short-term memory (LSTM) rekurentnoj mreži te dobivene rezultate usporediti s nekoliko radova koji su se bavili ovom temom. ; In this paper we discuss source code classification problem, the goal of which is to determine the functionality of a code snippet using programming language processing. We propose a solution in the form of a neural network based on the long short-term memory (LSTM) recurrent network and compare our results to several other papers that had discussed this topic. |
|---|