Abstract
dc:descriptionThe main goal of this work is to create a customer feedback system, which allows companies to look through, analyse and respond to the opinions that are formed about their product. Meanwhile the system also allows for clients to easily leave some feedback about the product and lets the user contact a feedback analyst. The positivity of opinions is predicted by using a trained multiclass classification model and the most negative feedback is highlighted in the system. To achieve this, the analysis of similar systems that are currently operating in Lithuania, as well as abroad, has been conducted. Machine learning algorithms and classifiers are examined, such as, max entropy, naive Bayes, decision trees. Best performer proved to be max entropy classifier. The work is done using C#, Angular, ASP.NET, ML.NET, EF technologies. The achieved result is a customer feedback system in which the feedback is evaluated by a trained multiclass classification model and presented to an analyst, who takes into account the negativity of the feedback and keeps contact with the client who raised it. System also contains regular users, who are able to create and amend feedback items. The are also administrator type users who not only have analyst privileges, but also have access to control of system users. System is developed using English language, as in order to train my model i need a lot of data, which can only be provided by international companies.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2020
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Veretenikas, Aleksandras,
- Contributors dc:contributor
-
- Misiukas Misiūnas, Andrius Vytautas
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD81917743&prefLang=en_US