Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: THEend8_
CCB302 – Digital Media Analytics Assessment 2: Social Media Analysis Brief on the Voice Referendum Document history CCB302 – Digital Media Analytics - semester 2 2023 – A2 Project Brief 2 Introduction CCB302 students are being asked to perform an analysis of real-world data collected over a period of time. The topic is the 2023 Voice Referendum. As part of this assignment, you are encouraged to review the upcoming Referendum in which all eligible Australian voters must vote on 14-Oct 2023.1 Note: For over a decade, the Voice Referendum has been supported by governments led by both Australian Coalition and Labor governments. More recently, the issue has become politicised by the Opposition and several prominent Aboriginal politicians. Many Australians consider the Referendum a human rights issue. You may find that contemporaneous issues such as an investigation of Australia’s national airline, Qantas, has been woven into the Voice Referendum discussion. This may inform part of your analysis. NOTE: This is not a hypothetical case. We are analysing real world data on the upcoming Voice Referendum. For this project, we are analysing real-word data that is deeply personal for many people, first & foremost, Australia’s First Peoples. The datasets contain emotional, and sometimes hyperbolic content. As digital communication students and soon communication professionals, this reflects the type of analysis one does in the ‘real world’ and it can be disturbing. If it raises any serious concerns, please contact QUT Student Counselling immediately.2 Background and methods In this exercise, you are provided four data sets with video and comments data, similar to Assignment 1; however, instead of a single video (on AI and healthcare), students will have four files to facilitate temporal analysis (before & after Garma). We will discuss how to do this in tutorials, or you are encouraged to experiment on your own using what you learned about JOINING datasets earlier this semester (for A1). 1 Aboriginal and Torres Strait Island Voice. The Voice Referendum will be held on 14-Oct 2023, and Australians will have a say about whether to change the Australian Constitution to recognise First
CCB302 – Digital Media Analytics - semester 2 2023 – A2 Project Brief 3 To help you get started, we have provided preliminary metrics on the YouTube data. As you begin your own analysis, you may ‘sanity check’ your initial findings in terms of number of unique videos, dates and channels posting content against our initial analysis. If there are discrepancies, seek to understand why and explain them. This is raw real-world data, so there will be inconsistencies and missing data (i.e., videos, comments, and users who are missing or have been deleted). That is usual and expected; you are welcome to identify that in your analysis. There are no intentional ‘red herrings’. We want you to explore the data with an open mind and surface insights. And remember, social media content is ephemeral! The data were retrieved from YouTube Data API using the QUT Digital Observatory’s tool call ‘Youte’. You have previously received information about this tool and encouraged to review the lectures and URLs provided to refresh your memory on how it works. You are being asked to perform mixed methods analyses using YouTube metadata. Textual analysis is done using Leximancer. Quantitative and qualitative data can be plotted using Tableau. You are welcome to incorporate other tools you may be familiar with, however, you must use Leximancer and Tableau at a minimum. The matching criteria was as follows: • “voice referendum australia” • published before 19 June 2023 • This date is before the 2023 Garma Festival of Traditional Cultures, Australia’s largest Indigenous cultural gathering taking place in early August in northeast Arnhem Land.3 These datasets are called “before Garma”. • We have collected data after the Garma Festival, between 1 and 15 August 2023, called “after Garma.” 3 The Garma Festival of Traditional Cultures (Garma) is Australia's largest Indigenous cultural gathering, taking place over four days each August in northeast Arnhem Land, in the Northern Territory, Australia. Hosted by the Yothu Yindi Foundation, Garma is a celebration of the cultural traditions of the Yolngu people, and a major community gathering for the clans and families of the Arnhem Land region. The event showcases traditional miny'tji (art), ancient story-telling, manikay (song), and bunggul (dance). It is held at Gulkula, a significant Gumatj ceremonial site about 40 kilometres (25 mi) from the township of Nhulunbuy, attracts more than 2500 guests each year and is often sold out months in advance. In recent years, Garma has become an important fixture on the political calendar, attracting business, political, academic, and philanthropic leaders to help shape Indigenous affairs policy through the Key Forum conference. Retrieved 09-09-2023 CCB302 – Digital Media Analytics - semester 2 2023 – A2 Project Brief 4 Preliminary Analysis - Before Garma Festival of Traditional Cultures Video stats before Garma • Number of videos: 351 • Earliest matching video: 2017-05-23 05:52:44 (this is an outlier that will need to be excluded) • Latest matching video: 2023-06-18 22:47:32 Video count by date CCB302 – Digital Media Analytics - semester 2 2023 – A2 Project Brief 5 CCB302 – Digital Media Analytics - semester 2 2023 – A2 Project Brief 6 Comments • Total number of comments collected: 28,435 Video stats after Garma Videos • Number of videos: 132 • Earliest matching video: 2023-08-01 04:57:59 • Latest matching video: 2023-08-14 15:10:49 Video count by date CCB302 – Digital Media Analytics - semester 2 2023 – A2 Project Brief 7 Comments • Total number of comments collected: 10,160 CCB302 – Digital Media Analytics - semester 2 2023 – A2 Project Brief 8 Suggested prompts for your analysis (qualitative and quantitative) 1 Who are the channels publishing videos before & after Garma on this topic? Were they the same or different? 2 You can hone in on the top 5-10 publishers and interrogate information on the leading publishers. 3 Alternatively, you may wish to examine the ‘long tail’ of publishers of content and discuss that. 4 What is their geographic location? (Australia, overseas?) What can you deduce from the level of interest delivered by platforms in languages other than English? 5 How many followers do they have? 6 What categories do they publish? 7 Are they considered an authoritative source? 8 Consider the [political] alignment (conservative to progressive). 9 What is their geographic reach? (What is YouTube’s geographic reach?) 10 What is the timeframe of the post (original post & comments)? 11 How many people commented (total vs. distinct)? 12 What can be said about the nature of comments based on the metrics? 13 Is there any spamming, trolling, bot activity, or any indication of automated posting? 14 What other events were co-incident with this time period? 15 What is the sentiment of the videos published by the top 5 – 10 publishers. What can be said about the nature of comments? 16 What themes or topics surfaced before and after the Garma Festival (using textual analysis approach)? How do you know? 17 If you ‘tune’ Lexicmancer, be sure to keep careful track of what changes you made and discuss them in the analysis. We will explore this in more detail during the remaining tutorials. 18 What tool(s) was used to retrieve YouTube metadata? Who produced it? (hint: see https://youte.readthedocs.io/en/latest/). 19 What are the rate limitations of the YouTube API quote system? CCB302 – Digital Media Analytics - semester 2 2023 – A2 Project Brief 9 21 Are there other tools you could have used for gathering video data and statistics? If so, what is another tool you identified? 22 What are some of the limitations of analysing metadata from digital media platforms generally, or YouTube specifically? 23 What additional data would have made your analysis more robust and thorough? 24 What are some of the data governance considerations 25 Are there any other things you can tell that we didn’t think of? The above are some of the prompts you may wish to investigate. There may be other insights you find and you are encouraged to discuss them! CCB302 – Digital Media Analytics - semester 2 2023 – A2 Project Brief 10 YouTube API Data Dictionary Video metadata Field Description kind Identifies the API resource’s type. The value will be youtube#video. id Unique identifier of the video, as provided by YouTube. published_at The date and time that the video was published (in UTC time). channel_id Unique identifier of the channel this video belongs to, as provided by Youtube. title Title of the video. description Full description of the video. thumbnail_url URL of the video’s thumbnail. thumbnail_width Width of the thumbnail image in pixels thumbnail_height Height of the thumbnail image in pixels. channel_tittle Title of the channel this video belongs to. tags A list of keyword tags associated with the video. category_id The YouTube video category associated with the video. localized_title The localised video title for a specified language. Returns the default language if no language is specified. localized_description The localised video description for a specified language. Returns the default language if no language is specified. default_language The language of the text in video’s title and description. default_audio_language The language spoken in the video’s default audio track. CCB302 – Digital Media Analytics - semester 2 2023 – A2 Project Brief 11 duration Length of the video in ISO 8601 duration (PT#M#S). The letters PT indicate that the value specifies a period of time, and the letters M and S refer to length in minutes and seconds, respectively. For example, a value of PT15M33S indicates that the video is 15 minutes and 33 seconds long. If the video is at least one hour long, the duration is in the format PT#H#M#S. dimension Indicates whether the video is available in 3D or in 2D. definition Indicates whether the video is available in high definition (HD) or only in standard definition. caption Indicates whether captions are available for the video. licensed_content Indicates whether the video represents licensed content, which means that the content was uploaded to a channel linked to a YouTube content partner and then claimed by that partner. projection Specifies the projection format of the video (either 360 or rectangular) upload_status The status of the uploaded video. privacy_status The video’s privacy status. license The video’s license (e.g. creativeCommon or YouTube). embeddable This value indicates whether the video can be embedded on another website. public_stats_viewable This value indicates whether the extended video statistics on the video’s watch page are publicly viewable. By default, those statistics are viewable, and statistics like a video’s viewcount and ratings will still be publicly visible even if this property’s value is set to false. view_count The number of times the video has been viewed. like_count The number of users who have indicated that they liked the video. comment_count The number of comments for the video. Will be empty if the video disables comments. topic_categories A list of Wikipedia URLs that provide a high-level CCB302 – Digital Media Analytics - semester 2 2023 – A2 Project Brief 12 description of the video’s content. live_streaming_* These values will only be available if the video is an upcoming, live, or completed live broadcast. meta_* Metadata of the youte version to collect data, time the data was collected. Comments metadata Field Description id Unique identifier of the comment, as provided by YouTube. video_id The ID of the video that the comment refers to.