Context
Original summaries
The paragraphs the 2023 dashboard showed beside each chart (written by the team).
page 1 of 1 CSV
| sourceTEXT | datasetTEXT | paragraphINTEGER | textTEXT |
|---|---|---|---|
| twitter_vic_sal_2022_02_2022_07 | 0 | Twitter data is in JSON format, a lightweight and widely used data interchange format. Within this format, each individual tweet is encapsulated as a JSON item, encompassing various key attributes. These attributes include the unique identifier of the tweet on Twitter (tid), the username of the author who posted the tweet (author), the precise timestamp indicating the date and time of its creation (date), the language in which the tweet is composed (lang), the actual textual content of the tweet itself, location of where this tweet is made (location), the corresponding SAL code denoting the specific region (sal), and finally, a sentiment score ranging from 1 to 9. This sentiment score serves as an evaluative measure, providing insights into the sentiment conveyed by the tweet, with lower scores reflecting more negative sentiment and higher scores indicating a more positive sentiment. | |
| twitter_vic_sal_2022_02_2022_07 | 1 | Based on the analysis of Twitter data in Victoria, it was found that the distribution of tweets mentioning income-related terms is highly uneven across SA2 regions. Approximately 73% of the regions have only 1-10 tweets related to income, while 22% have 10-100 tweets, and only 4.7% have more than 100 tweets. Notably, Melbourne stands out with a significant concentration of income-related tweets, totaling 23.281k tweets, surpassing other regions by a large margin. However, the SUDO data, which indicates median income, does not show a similar concentration in high-income regions around Melbourne. | |
| twitter_vic_sal_2022_02_2022_07 | 2 | The average sentiment scores of tweets discussing income-related topics exhibit an even distribution, ranging from 1 to 9, without any clear clusters or patterns. Therefore, there is no clear indication that people in higher or lower median income areas tweet more frequently about income-related topics, and no significant correlation between sentiment and income-related tweets was found. | |
| twitter_vic_sal_2022_02_2022_07 | 3 | Regarding crime-related topics, no significant clusters or patterns were identified in the average sentiment scores of tweets discussing crime. However, in the second scenario, it was observed that Ballarat, with a relatively high occurrence of offenses and a significant number of tweets mentioning crime-related terms, has a higher probability of being a dangerous region. Nonetheless, there is no significant correlation between the sentiment of tweets and crime-related topics. | |
| twitter_vic_sal_2022_02_2022_07 | 4 | In summary, the analysis of Twitter data does not provide strong evidence to suggest that people in higher or lower median income areas tweet more about income-related topics or that sentiment of tweets is correlated with income. Similarly, no significant correlation between sentiment and crime-related tweets was found, although Ballarat stands out as a region with a higher probability of being dangerous based on crime occurrence and tweets mentioning crime-related terms. | |
| sudo | median_income_sa2 | 0 | The income dataset from sudo provides insights into income distribution across Australia for the financial year of 2015-2016. It is based on the Statistical Area Level 2 (SA2) regions according to the 2016 edition of the Australian Statistical Geography Standard. The analysis focuses on the mean, median, sum, and median age of earners, using the SA2 code 2016 column to join with the shapefile for visualization purposes. |
| sudo | median_income_sa2 | 1 | Out of the original 2288 SA2 regions, 2234 remain after removing entries with null values. The dataset is then grouped by Greater Capital City Statistical Areas (GCCSAs) to match the granularity of the available Twitter data. The summary statistics reveal no outliers in the income mean, median, and sum columns, so no outlier removal is necessary. |
| sudo | median_income_sa2 | 2 | Upon examination, the Rest of West Australia (5RWAU) region exhibits the highest income mean of 71.5k AUD, while the Rest of Victoria (2RVIC) has the lowest income mean at 49.6k AUD. This indicates a higher average income in the Northwest compared to the Southeast. The Australian Capital Territory (8ACTE) boasts the highest median income of 60.2k AUD, which is expected given that the capital, Canberra, is located there. In contrast, the Rest of Victoria (2RVIC) records the lowest median income of 41.4k AUD. Greater capital cities generally have a median income of around 50k AUD. |
| sudo | median_income_sa2 | 3 | The sum of income shows an opposite trend to the average income, with eastern regions being higher than western regions. This suggests a higher population in the eastern regions compared to the western ones. As for the median age of earners, the regions are quite similar, with the exception of the Australian Capital Territory (8ACTE), which appears to have a younger earning population. |
| sudo | crime_vic_lga | 0 | The crime dataset from sudo provides a comprehensive overview of criminal incidents by principal offense recorded on the Victoria Police Law Enforcement Assistance Program (LEAP) and is aggregated to the 2011 Australian Statistical Geography Standard (ASGS) Local Government Areas (LGA). Analysis is focused on the total number of all kinds of offences, using the LGA code 2011 for spatial representation. |
| sudo | crime_vic_lga | 1 | Upon examining the data, outliers were identified and subsequently removed for a more accurate analysis. The total number of offences for each category was calculated, and a summary table was generated. |
| sudo | crime_vic_lga | 2 | The distribution of different types of offences across the region was found to be quite similar. Brimbank had the highest number of cases among all types of offences. Several suburbs near the Central Business District (CBD) such as Wyndham, Melton, and Whittlesea showed relatively high crime rates; however, this could be attributed to their larger population sizes. Apart from these suburbs near the CBD, other areas with high numbers of offences included Mildura, Ballarat, Greater Bendigo, Greater Shepparton, Latrobe, Frankston, and Mornington Peninsula. These areas may be considered more dangerous districts due to their elevated crime rates. |
| mastodon | mastodon | 0 | Three Mastodon servers are utilized: Mastodon AU, Mastodon Social, and Mastodon TikTok. |
| mastodon | mastodon | 1 | Mastodon AU instances specifically cater to the Australian user base, providing a platform for Australians to connect, share content, and engage in discussions. Each Mastodon AU instance may have its own rules, moderation policies, and community guidelines, fostering a diverse and localized online community experience for Australian users. |
| mastodon | mastodon | 2 | Mastodon Social is one of the largest and most popular instances on the Mastodon network. It is known for its diverse user base and active community. It offers features like customizable profiles, timeline filtering, and private messaging, while adhering to Mastodon's decentralized and federated structure. |
| mastodon | mastodon | 3 | Mastodon TikTok allows users to engage in decentralized social networking, share content, and connect with other users within the instance. |
| mastodon | mastodon | 4 | The data structure of each toot in all three servers is identical, comprising a toot ID (TID), a precise timestamp indicating the date and time of creation (date), the username of the author (author), the language of the toot (lang), the content of the toot (content), and a sentiment score for each toot (score). Unlike Twitter data, Mastodon lacks geoinformation and does not provide any indication of the location where the toot was created. |