diplomsko delo
Abstract
Nezaželena akademska elektronska sporočila so nezaželena sporočila, ki jih prejemajo profesorji, raziskovalci in drugi akademiki. Njihov namen je priti do zaslužka z zahtevanjem plačila za objavo prejemnikovega članka v reviji, s promoviranjem komercialnih revij ali plačljivih konferenc. Neizkušenemu prejemniku se lahko takšna sporočila sicer zdijo dobronamerna in laskava, vendar pa odgovarjanje nanje lahko škoduje prejemnikovi karieri.
V okviru diplomske naloge smo izdelali filter nezaželene akademske elektronske pošte za Gmail, ki izmed neprebranih elektronskih sporočil v uporabnikovem elektronskem nabiralniku označi takšna nezaželena sporočila. Preizkusili smo več klasifikacijskih modelov, v končnem sistemu pa smo uporabili najuspešnejšega, in sicer nevronsko mrežo v kombinaciji z vektorsko vložitvijo besed. Izdelan sistem ima možnost posodobitve klasifikacijskega modela glede na nezaželeno akademsko elektronsko pošto, ki jo je uporabnik prejel v preteklosti. Uspešnost modela smo primerjali z obstoječimi metodami klasifikacije nezaželene elektronske pošte in potrdili, da je izdelan model primerljiv ali celo boljši od obstoječih metod.
Keywords
nezaželena elektronska pošta;klasifikacija;filter;akademska sporočila;računalništvo in informatika;univerzitetni študij;diplomske naloge;
Data
Language: |
Slovenian |
Year of publishing: |
2022 |
Typology: |
2.11 - Undergraduate Thesis |
Organization: |
UL FRI - Faculty of Computer and Information Science |
Publisher: |
[A. Vrečer] |
UDC: |
004.773.3(043.2) |
COBISS: |
99811075
|
Views: |
159 |
Downloads: |
28 |
Average score: |
0 (0 votes) |
Metadata: |
|
Other data
Secondary language: |
English |
Secondary title: |
Academic spam filter |
Secondary abstract: |
Academic spam messages are unsolicited e-mails that professors, researchers, and other academics receive in their inboxes. Their purpose is to make money by charging fees to submit the receiver’s article in their journal, to promote commercial journals or payable conferences. An inexperienced receiver might think that messages of this kind are well-intended and flattering when in reality answering them could have devastating consequences in the receiver’s career. In this thesis, we developed an academic spam filter for Gmail that labels academic spam messages in the user’s inbox. We tested different classification models for classifying academic spam and used neural networks in combination with word to vector embedding in our final system since it has shown to be the most effective.Users can also update the classifier in our system based on the academic spam messages that they have received in the past. We compared our model with the existing methods of spam classification and confirmed that it is comparable or even better than them. |
Secondary keywords: |
spam messages;classification;filter;academic spam;computer and information science;diploma;Elektronska pošta;Računalništvo;Univerzitetna in visokošolska dela; |
Type (COBISS): |
Bachelor thesis/paper |
Study programme: |
1000468 |
Embargo end date (OpenAIRE): |
1970-01-01 |
Thesis comment: |
Univ. v Ljubljani, Fak. za računalništvo in informatiko |
Pages: |
52 str. |
ID: |
14673459 |