Please use this identifier to cite or link to this item:
http://dspace.cityu.edu.hk/handle/2031/9554
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | De, Ankolika | en_US |
dc.date.accessioned | 2023-03-15T10:12:06Z | - |
dc.date.available | 2023-03-15T10:12:06Z | - |
dc.date.issued | 2022 | en_US |
dc.identifier.other | 2022csda200 | en_US |
dc.identifier.uri | http://dspace.cityu.edu.hk/handle/2031/9554 | - |
dc.description.abstract | Instagram is a social networking service (SNS), used predominantly on mobile devices for the purpose of sharing pictures with captions. Lately, it has been used as more than a personal photo sharing application, as users appropriated it to form communities that share, and encourage sharing of different kinds of content. The poetry community of Instagram is one such community which shares poetry with pictures or graphics on the application for multiple reasons. This community is interesting because they have managed to appropriate a predominantly visual social media with recommendation algorithms as a poetry sharing medium, sharing linguistically heavy content as compared to other posts. Poetry has often been looked into as a traditional art form, and using a digital space to create and share it is novel. Very little is known about the poetry community from an academic perspective. Communities like these have been looked into in the past, as they are useful for studying communication, and understanding youth culture, misinformation and recommendation algorithms. This study, attempted an empirical, interview-based study of this community, to note their motivations, challenges and perceptions of the Instagram algorithm from a Human-Computer-Interaction (HCI) perspective. The implications of the findings were discussed. Discussions regarding the major motivations indicated that the main reasons for using Instagram to share poetry was ease of usage and instant feedback sharing opportunities, with "comments" being the most popular form of metric. All interviewees believed that the Instagram algorithm lacked transparency and disadvantaged them by favoring reels and videos over content requiring more attention to engage with. This frustrated the interviewees, and also discouraged many to discontinue posting or writing poetry. The discussions also brought in challenges, among which the biggest problem was about having to create the right visual content, without knowing any measures to increase their engagement. The interviewees were unhappy that even though they tried very hard to make engaging visuals, their posts often did poorly and it was puzzling to understand what went wrong and understand the recommendation algorithm. They further spoke about how they follow verified poetry accounts (accounts authenticated by Instagram to represent a brand or celebrity represented with a blue tick). However, it was hard for them to understand the metrics that helped these poets gain the reach that they had with the current recommendation system. As a result of this, a creativity support tool (CST), was built that could help novice poets understand which type of visuals are best suited for their poetry. A comprehensive literature review of prior CST(s) that were built for novices was conducted, and a list of design goals that was focused on novices was made from contributions from prior scholarship, as well as findings from the formative study done by the author. As this area is unique, the design was highly empirical. The tool was made in the form of a website that helps poets get inspiring visuals which performed well on Instagram and was similar to their content in terms of ideas, themes and content. The main purpose was to provide assistance in the creative pipeline, where the poets were open for intervention, and as they found it difficult to determine what kind of visuals would engage better and were open to changing and designing it, the "visual creation area" was chosen to be the area that the author could provide technological assistance in. The theme and building of this CST was derived from multiple Computer-supported cooperative work (CSCW) themed research papers. The final design consisted of having a database that had the posts and comment files of the poets who were verified on Instagram, and this was used to suggest similar and good performing posts to the novice poets. The website had a ReactJS frontend deployed on Netlify and a Python backend deployed on a Heroku server. The backend used various algorithms from Computer Vision (CV), and Natural Language Processing (NLP) such as the text extraction algorithm provided by pytesseract to extract text from images, sentiment analysis to calculate polarity of instagram comments in posts, and cosine similarity to match sentences and calculate similarity. Being a tool that is primarily used for creative design assistance, it followed agile software development practices, with constant user interaction, feedback and planning. A carefully designed user study validated the design goals and provided satisfactory results. The frontend design was made simple, sustainable, scalable and intuitive to provide users with the best experience after having many face-to-face feedback sessions. The system was tested using snapshot testing, to ensure that the UI, which is mostly static, does not change in an unexpected or unstable way. Finally, the discussions from the findings and the user study, revealed some key understandings regarding the community and future areas that needed further research. | en_US |
dc.rights | This work is protected by copyright. Reproduction or distribution of the work in any format is prohibited without written permission of the copyright owner. | en_US |
dc.rights | Access is restricted to CityU users. | en_US |
dc.title | Understanding the Practices and Challenges of Poets on Instagram and Designing a Creativity Support Tool to Assist them with Content Creation | en_US |
dc.contributor.department | Department of Computer Science | en_US |
dc.description.supervisor | Supervisor: Dr. Lu, Zhicong; First Reader: Dr. Wang, Shiqi; Second Reader: Prof. Liang, Weifa | en_US |
Appears in Collections: | Computer Science - Undergraduate Final Year Projects |
Files in This Item:
File | Size | Format | |
---|---|---|---|
fulltext.html | 147 B | HTML | View/Open |
Items in Digital CityU Collections are protected by copyright, with all rights reserved, unless otherwise indicated.