Filtering by
- All Subjects: Biclustering
- All Subjects: Coliving
- Creators: Chavez-Echeagaray, Maria Elena
Coliving is a concept that has many benefits towards society and sustainability. This is due to the resources saved economically and environmentally when living with other people. Aisha Comfortable Coliving, a company based in Canada, provides a service where they help women find Coliving communities. A lack of knowledge pertaining to this service could slow down or halt the growth of Aisha ElSherbiny’s Aisha Comfortable Coliving company. This thesis was an extension of a broader project, “Web App for Aisha Comfortable Coliving Inc.,” which focused on transitioning from their current website platform into a web application. As an extension of this main project, this thesis is focused on the engine component design portion surrounding AI chatbots to determine which implementation would provide the best results for a small company in reaching their target audience and helping inform them through an interactive chatbot. The ability to present 24/7 support for Aisha Comfortable Coliving brings value to the company and the methods used in this chatbot can be reproduced in order to create similarly effective chatbots. This thesis delves into the various approaches and implementations researched to determine how to optimize the backend of a chatbot to provide speed, reliability, and expandability for companies aiming to create a chatbot for their users to interact with. It also discusses the methods used when implementing a chatbot called AishaBot using the IBM Watson Assistant’s platform that includes the development of Intents, Entities, Dialog Tree structure, and its WebHook functions. Overall, satisfaction pertaining to the designed chatbot engine within IBM Watson Assistant was discovered to be positive through user trials. Limitations have been discovered, feedback for future improvements have been noted, and lessons learned about the thoroughness of training data have been discussed.
Moreover, most of these recommender systems suffer from the cold-start problems where insufficient data for new users or products results in reduced overall recommendation output. In the current study, we have built a recommender system to recommend movies to users. Biclustering algorithm is used to cluster the users and movies simultaneously at the beginning to generate explainable recommendations, and these biclusters are used to form a gridworld where Q-Learning is used to learn the policy to traverse through the grid. The reward function uses the Jaccard Index, which is a measure of common users between two biclusters. Demographic details of new users are used to generate recommendations that solve the cold-start problem too.
Lastly, the implemented algorithm is examined with a real-world dataset against the widely used recommendation algorithm and the performance for the cold-start cases.