Model za preporučivanje citata za patente baziran na sadržaju te dubokom učenju

undergraduate thesis

• Title (english): Deep learning content-based citation recommender model for patents
• Author: Dora Šokota
• Mentor: Mario Brčić (mentor)
• Committee member: Mario Brčić (predsjednik povjerenstva)
• Committee member: Agneza Krajna (član povjerenstva)
• Committee member: Kristijan Poje (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-07-04, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Patent je isključivo pravo priznato za izum koji nudi novo rješenje nekog problema. U sklopu prijave potrebno je citirati patente da bi se specificiralo postojeće znanje na kojem je novi izum baziran. Godišnje broj patenata poraste za više od 150000 što otežava prihvat i obradu infromacija o njima. Kako bi se olakšalo korisniku da iz velike količine podataka izvuće podskup koji je njemu značajan, razvijen je algoritam za preporuku citata. U radu je opisana implementacija model baziranog na dubokom učenju koji temeljem sažetka izdvaja patente koji su prikladni za citiranje. Također, se radi zaštite izvornosti izuma korisnika, sažetak patenta se pohranjuje na javni blok-lanac Solanu. Korisnik kao potvrdu dobiva kriptografski sažetak adrese pohranjenog patenta.

Abstract (english)

A patent is an exclusive right granted for an invention that offers a new solution to a problem. As part of the application, it is necessary to cite patents in order to specify the existing knowledge on which the new invention is based. The number of patents is growing by more than 150,000 annually, which makes it difficult to receive and process information about them. In order to make it easier for the user to extract a subset that is important to him from a large amount of data, an algorithm for recommending citations has been developed. The paper describes the implementation of a model based on deep learning, which, based on the summary, singles out patents that are suitable for citation. Also, in order to protect the authenticity of the user's invention, the abstract of the patent is stored on the public blockchain Solana. The user receives a cryptographic summary of the address of the stored patent as confirmation.

• Keywords: patent; citati; izvornost podataka; bloklanac; transformer; obrada prirodnog jezika
• Keywords (english): patent; citation; data provenance; blockchain; transformer; natural language processing
• Language: croatian
• URN:NBN: urn:nbn:hr:168:604054
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-15 11:20:41