Summary Over the past two semesters, IRM capstone team DJKYD (amalgamation of our names) identified an information management problem and embarked on a process to research, design, implement and present a solution in the form of a Carleton University artificial intelligence (AI) support hub website (also viewable in this demo video). At the very beginning in the ideation stage, we recognized how impactful AI has become in the past few years, especially in the academic setting. In conjunction, given this uptake, we found that Carleton doesn’t have one online space for providing support in this specific area. Uncertainty is a feeling humans are known to strongly dislike, though it has continued to trickle into many of us when thinking about the scalability of this technology, faculty addressing AI in the classroom, students completing coursework and careers that might be stopped short from AI takeover. While the increasing influence and capabilities of AI is inevitable, Carleton and other academic institutions should be central disseminators of support and knowledge as they prepare the future workforce. Given the accessibility of AI and its certain existence in the future, people should be taught how to utilize it appropriately, otherwise they are more likely to exploit it. Fully discouraging AI use is not realistic nor is necessarily productive when we can make our lives more efficient in some aspects. Our simplified motivation statement is as follows: “Due to the significant advancement of AI, we want to support the Carleton community in navigating coursework and career paths, all in one place”.
Since our previous report that presented our design process, we have fully implemented and completed our Capstone product. From this, we have all understood what it takes to complete a larger scale project from start to finish, and how directions often diverge since its beginning. Our hope is that this finished capstone product and all our documented processes acts as a starting point for Carleton in uplifting the community it serves.

Summary The Carleton Clinic currently lacks a dedicated website, making accessibility and navigation difficult, and thus hindering its ability to provide essential services and information to users. Notably, users are unable to natively book appointments online and must do so via phone or a third party booking system that holds different data privacy and usage standards than Carleton. Also, the existing web pages suffer from lackluster user flows and features, making navigation, accessibility, and the overall user experience subpar. Additionally, the absence of a clearly defined privacy policy, patient data usage policy, and other similar information raises concerns about transparency and data security. These issues collectively create a suboptimal experience for patients and highlight the need for a comprehensive, user-friendly, and secure online platform.
Project Goal: Securing Health: Bringing improved data security, fresh features, and user accessibility to the Carleton Clinic
Project Objectives: To achieve our project goal, our team developed a new dedicated website for the clinic that had an emphasis on enhancing accessibility, navigation, security, transparency, and the overall user experience. To ensure this mandate, we included an integrated online booking system that leverages blockchain technologies to improve encryption and digital security. Additionally, to improve transparency and user awareness, we created an informative policies page that summarizes the Carleton Clinic’s various privacy, data usage, and other key policies in plain language. Finally, we aimed to improve the user experience by adding new features like a calendar scheduling tool and an AI assistant.
Project Novelty:

• An in application booking system using blockchain encryption technology to better gather, store, share and safeguard user privacy.
  • Automated reminders and notifications.
  • User authentication.
• Informative policies page of Carleton Clinic's privacy policy, patient confidentiality policies, terms of service, etc. Supports transparency and raises user awareness.
• An integrated calendar scheduling tool that shows available appointment times.
• An in application AI chatbot for common navigation questions regarding our application.

Summary Managing and cataloging large volumes of records is a persistent challenge in libraries and archives. Existing tools often require specialized knowledge that can be difficult for novices to navigate and completely impossible for someone without any official training. PallasOwl was made to address these challenges by providing a tool that simplifies cataloging tasks, ensures adherence to MARC21 metadata standards, and offers a wide variety of features to accommodate both professionals, novices, and lay-people. This report outlines the problem in cataloging, the solution provided by PallasOwl, and its innovative approach to improving the cataloging process.
Problem Statement Within the cataloging field, librarians have lacked the necessary tools needed to manage the ever-growing volume of records that they have accumulated over the years. The challenge lies not just in managing these vast collections, but in making sure everything stays consistent and adheres to standard practices. Existing tools struggle to process the massive scale of records and are not able to properly identify inconsistencies across them. Furthermore, the creation and maintenance of cataloging records often requires specialized knowledge that is tailored towards experienced professionals. As a result novices may struggle with using these tools due to a limited understanding of how they function.
Solution PallasOwl uses tools that simplify tasks which traditionally relied on manual entries. They help librarians maintain adherence to established metadata standards. PallasOwl provides users with both commonly used entries, but also the ability to add more customization through the advanced panel. This ensures that users have the ability to create and modify MARC fields they’re interested in. The tool also provides those with limited experience creating mrc files the ability to perform cataloging at a pace they are able to understand. Offline-accessible help buttons and support for multiple subject headings make the cataloging process more approachable for beginners, while still providing the flexibility needed by professionals. The ai features also provide guidance for DDC and LCCS suggestions, which are able to identify suggestions for improvement. In addition, PallasOwl provides a search database that allows catalogers to view their records. In addition to being able to search for their records, users may also visualize the data. This ensures that users are able to identify common trends or potential outliers within their collection

Project Summary When describing resources, folksonomic tagging refers to the assignment of descriptive labels outside the control of an authority or controlled vocabulary. Instead, folksonomies are controlled by the users (e.g., library patrons) themselves. Folksonomies can significantly enhance the searchability and discoverability of a library’s catalogue, but they also face fundamental issues related to synonymy, homonymy, polysemy, and plurality. Currently, websites with the best folksonomic tagging systems use labour-intensive tag-wrangling strategies to resolve these problems, which is impractical for libraries operating on limited budgets and resources. Moreover, research has shown that engagement with library tagging systems has been notoriously low.
Therefore, we have developed a website to showcase a new design solution. This solution introduces gamification as a method for boosting user engagement with libraries’ tagging systems. It also proposes an innovative, automated solution to the problems of synonymy and plurality via the use of word-finding APIs and natural language processing tools such as textblob. The website has been carefully built with all basic user requirements in mind, including user accounts, searching and browsing, tag management, community-based tag curation, user comments, and mobile support. All website features function as intended, and all pages respond in less than a second—with the exception of the Search Results page, which is dependent on the Library of Congress’ own response times. Furthermore, our website successfully fulfills several user experience needs that other folksonomic tagging websites lack in some way or another. Finally, our solution simplifies many user workflows by integrating and automating various complex processes such as tag wrangling and conflict verification.