1103 ASSIGNMENT BIG DATA ANALYSIS PDF

Preview

Summary

This is an assignment for Big Data analysis for you tube uploaded videos for 2006 -2018 and it cover all the dashboards and recommendation for the content manager....

Description

BIG DATA ANDANALYTICS ANALYTIC REPORT

MANJOT KAUR(30356880) PRABHJOT KAUR(30357945) HIMANSHU(30359383)

ITECH 1103

Table of Contents 1. Executive Summary..........................................................................................................................3 2. Introduction....................................................................................................................................4 3. Task 1: Project Background...............................................................................................................5 4. Task 2: Reporting/Dashboards.........................................................................................................7 4.1 Number of unique videos, categories, countries and channels (1-4).............................................7 4.2 Top three and least three countries in terms of channels including US (5-7)..............................7 4.3 Top and least 10 viewed video titles across different countries (8-9)............................................8 4.4 Total number of years, videos in last month and yearly count of videos uploaded in GB (10-12)..9 4.5 Most uploaded videos by hour, top/least 3 viewed categories (13-15)...........................................9 4.6 Videos with highest likes and dislikes, day with highest and least uploads (16-19)....................10 4.7 Monthly breakdown of published videos (20)...........................................................................11 5. Task 3: Advanced Insights..............................................................................................................12 5.1 Insight 1.....................................................................................................................................12 5.2 Insight 2...................................................................................................................................12 5.3 Insight 3...................................................................................................................................12 5.4 Insight 4...................................................................................................................................12 5.5 Insight 5...................................................................................................................................12 6. Task 4: Research............................................................................................................................13 7. Task 5: Recommendations for Content Manager............................................................................15 8. Task 6: Cover letter........................................................................................................................16

9. Task 7: Reflection...........................................................................................................................17 9.1 Reflection by team member 1....................................................................................................17 9.2 Reflection by team member 2...................................................................................................17 9.3 Reflection by team member 3...................................................................................................17 10. Conclusion...................................................................................................................................18 11. References.....................................................................................................................................19

1. Executive Summary This report has been developed as a Content Analyst in an ABC online multimedia company. The aim of this report is to presents the findings of analytical project commissioned by Content Manager of ABC online multimedia that are derived from visualisation solutions. IBM Watson Analytics was applied to design 20 different visualisation solutions for the analytics problems and subsequently business intelligence (BI) dashboards were designed to report the useful insights, pattern and trends identified in the YouTube dataset. The most basic findings were that 161,470 videos were uploaded by 12,360 channels. These videos were uploaded across 18 different YouTube categories whereas top three categories identified with category ids are (20) people and blogs, (24) entertainment, and (25) news and politics. (30) Movies, (43) shows, and (44) trailers are the three categories of YouTube that are least watched by YouTube audience. The scope of this analytical report is to provide background information on the YouTube dataset, report the insights, patterns and trends through visualised business intelligence (BI) dashboards with application of IBM Watson Analytics. The scope of this report is to present additional five insights to help content manager make suitable decisions regarding advertisements and content over which advertisements should be placed and justification for the selection of BI reporting dashboards will be given with support of researched evidence. This report will provide recommendations for content manager to improve multimedia operations and justifications will be provided as to the way different recommendations can improve multimedia operations and help ABC in achievement of strategic objectives. Therefore, cover letter and three reflections are included.

2. Introduction The aim of this data analytical report is to provide insights, pattern and trends in YouTube dataset by applying IBM Watson Analytics to design visualisations and BI dashboards. The visualisations will be integrated within a dashboard to provide content manager of ABC with holistic insights about different videos that were uploaded on YouTube from 2006 to 2018. The report will identify total number of uploaded videos, different types of uploaded categories, number of countries, unique channels and top three as well as least three countries identified as per number channels. This report will also provide background information to ensure that the dashboards can be better understood, and advanced insights will be provided along with research to make implementable recommendations for the content manager. Cover letter will be developed outlining important data insights and recommendations for the achievement of company objectives. This report will complete the analytical project by recording the challenges, lessons learnt and contribution through individual reflection by each team member.

3. Task 1: Project Background This analytical report completes the analytical project given by content manager of ABC by using the given dataset on YouTube and applying IBM Watson Analytics to design and visualise useful insights, pattern and trends in the dataset. About the Firm:

You Tube,

is an American video-sharing website headquartered in San Bruno,

California. Three former PayPal employees—Chad Hurley, Steve Chen, and Jawed Karim—created the service in February 2005. Available content includes video clips, TV

show clips, music

,videos

, short and documentary,

films,

audio

recordings, movie trailers live streams, and other content such as video blogging, short original videos, and educational videos. Most of the content on YouTube is uploaded by individuals, but media corporations including CBS, the BBC, Vevo, and Hulu offer some of their material via YouTube as part of the YouTube partnership program. The project is based on using IBM Watson Analytics to generate visualisations for completion of the twenty different questions asked by the content manager. Furthermore, visualisations will be used to design BI dashboards that can provide a consolidated view of data. Hence, the dataset will provide information about the performance of different videos that were uploaded between 2006 and 2018. Benefits to You Tube The selected dataset of YouTube is important for this analytical project as it provides a big data set that requires application of the big data analytics. Therefore, this dataset provided an

opportunity for ABC to apply IBM Watson Analytics with the help of which analytical problems can be resolved and organisational objectives can be supported. Recent studies on visualised business intelligence reporting have informed that analysing the YouTube data is important for the multimedia companies to ensure that the high performing videos are targeted to place the advertisements and content as well as advertisements are produced by ABC for the highly viewed categories. The insights and findings arising from the application of IBM Watson Analytics will provide a direction to the content manager for the development of advertising, consumer and informational content. Moreover, this project is important for not only the content manager, but also the stakeholders and top management team of ABC to participate in the knowledge transfer and develop understanding about the viewing preferences related to categories and type of content among the YouTube audiences.

4. Task 2: Reporting/Dashboards 4.1 Number of unique videos, categories, countries and channels (1-4) The application of IBM Watson Analytics identified that there are 161,470 unique videos in the dataset. It was further found that dataset is reported for 18 unique categories. The dataset provided details about 161,470 videos from four countries, US, Canada, GB and France. The visualisation identified that there are 12,360 channels from the combination of four identified countries.

Dashboard 1: Visualised reporting for question 1 to 4

4.2 Top three and least three countries in terms of channels including US (5-7) Total number of unique channels in the US will be identified. The top three countries as per the number of channels are France at 6,679 channels followed by Canada at 5,076 channels, and US at 2,207 channels. GB had the lowest channels with the total count being 1,624. Three countries with the least channels are GB, US and Canada.

Dashboard 2: Visualised reporting for question 5 to 7 4.3 Top and least 10 viewed video titles across different countries (8-9) List of top 10 viewed and least 10 viewed video titles will be developed for each country. Total number of years, number of uploaded videos last months, and other important insights will be informed.

Dashboard 3: Visualised reporting for question 8 and 9 The top 10 viewed and least 10 viewed titles are listed in the dashboard above.

4.4 Total number of years, videos in last month and yearly count of videos uploaded in GB (10-12) 13 years of published data was analysed in the data set. IBM Watson Analytics suggested that 6,231 videos were uploaded last month and that most of the videos in GB were uploaded in 2018.

Dashboard 4: Visualised reporting for question 10 to 12 4.5 Most uploaded videos by hour, top/least 3 viewed categories (13-15) Most videos were uploaded during the time frame of 16:00 to 16:59. It was identified that most of the videos were uploaded by France followed by Canada, US and GB. Top three categories identified with category ids are (20) people and blogs, (24) entertainment, and (25) news and politics. (30) Movies, (43) shows, and (44) trailers are the three categories of YouTube that are least watched by YouTube audience.

Dashboard 5: Visualised reporting for question 13 to 15 4.6 Videos with highest likes and dislikes, day with highest and least uploads (16-19) The videos with highest percentage of likes was titled as “This is America” and it was published by Childish Gambino with 265 million likes. The most disliked videos were titled as “So Sorry” with 1.5 million dislikes. Big data analytics further helped in identifying the trend that most of the videos are published on Friday whereas least number of videos are published on Saturday.

Dashboard 6: Visualised reporting for question 16 to 19

4.7 Monthly breakdown of published videos (20) Monthly breakdown of videos indicates that GB published more videos that channels in France and that number of videos published by Canada was lowest. It was also recognised from the monthly trend analysis that lowest number of videos are uploaded in the months between July and October.

Dashboard 6: Visualised reporting for question 20

5. Task 3: Advanced Insights 5.1 Insight 1 One of the advanced insights obtained from the dataset is that while France has the highest number of channels, the videos uploaded by the French channels are equivalent to the number of videos uploaded by GB and US. 5.2 Insight 2 Second insight for content manager is that GB accommodates largest number of YouTube audience among the other channels. Top 10 videos on YouTube obtained highest viewed from only two markets of GB and US. 5.3 Insight 3 It was further identified that while most of the videos were uploaded by GB in 2018, the data for months between July and October shows lowest levels of video upload activity. It was also found that new channels can improve their visibility during July and October due to low upload activity. 5.4 Insight 4 Videos that were identified to be the least 10 viewed on YouTube were from France. This develops a risk that the channel from France might not perform as well as compared to channels from other countries. Another finding is that most of the videos that performed poorly in terms of views were uploaded in shows category. 5.5 Insight 5 One of the most important insight suggests that highest number of videos uploaded in GB are during 5 to 5:59 am on Friday as illustrated below. However, most of videos uploaded across Canada, France and US takes place during 16:00 to 16:59.

Dashboard 7: Country wise difference in timeframe of video 6. Task 4: Research Study by Bello-Organ et al., (2016) brought it to the attention that business intelligence (BI) dashboards can encourage the multimedia companies such as ABC to learn and design knowledge patterns across the target audience (Bello-Orgaz et al., 2016). The visualisation enables large datasets such as the YouTube dataset in this project to be analysed and mined for identification of patterns, trends and consumer preferences (Behera et al., 2017). Study by Hartmann et al., (2016) further recognised that the big data captures the insight from large dataset due to which the business organisations that uses data-driven business models are likely to achieve a competitive advantage over the other business players in the same industry (Hartmann et al., 2016). Journal article published by Yaqoob et al., (2016) was consistent with the views of other studies that the data visualisation enables the organisation to study the preferences of consumers by determining the areas where most of the consumers are engaged (Yaqoob et al., 2016). Hence, in the context of this project, it was recognised that not only primary target audience on YouTube is situated in GB, but also the interest of the target audience lies within the entertainment category of videos (Ghosh, 2018).

The BI reporting dashboards were laid out in the combination of two, three and four visualisations clubbed together within a single dashboard (Ye & Morro, 2018). Combination of different visualisations combined in one dashboard can be justified with the support of Sirin and Karacan (2017) as these researchers suggested that by combining the related information together, business executives and managers such as content manager can be supported in making decisions based on data and help in reducing the time required for decision making (Sirin & Karacan, 2017).

Screen Capture 1: Snip of dashboard 2 The screen capture above illustrates that the information related to number of channels in country will help content manager to determine the importance of each geographical location in dataset based on decision (Wamba et al., 2015). While the French market will be most competitive due to the number of channels, GB market will be most attractive due to the low number of channels than US and higher number of viewers in GB (Kline & Stokes, 2017).

7. Task 5: Recommendations for Content Manager This report provides following recommendations for content manager along with the individual justification for each recommendation: 1. It is to recommend the content manager of ABC that IBM Watson Analytics is used on regular basis to adopt a data driven business model wherein the decisions made by managers and top management team is based on the visualised data (Pouyanfar et al., 2018). This recommendation can be justified as IBM Watson Analytics can be used as a business intelligence tool through which visualised dashboard can be developed to aid and support content manager in decision making (Guidi et al., 2016). 2. This report further recommends content manager of ABC to place the advertisement on the videos that are published in the entertainment category of YouTube (Satish & Yusof, 2017). This recommendation can be justified as the entertainment category was identified to be the most viewed category on YouTube (Jha et al., 2016). Therefore, the advertisement on entertainment videos will be exposed to higher number of unique audiences (Sun et al., 2017). 3. It is also recommended that the advertisement and video content produced by ABC are promoted in GB to reach the maximum number of viewers and views (Khosla, 2016). This recommendation can be justified as the top 10 videos showed that most of the views were obtained from GB followed by US.

8. Task 6: Cover letter To: Content Manager YOUTUBE From: Content Analyst Subject: Findings from IBM Watson analytical project This letter has been drafted upon the completion of analytic project that was intended to explore, analyse and mine the data from YouTube dataset. IBM Watson Analytics was applied to design visualisations and BI dashboards that can provide useful insights, pattern and trends pertaining to the dataset. Visualised reporting dashboards were designed and included within the analytical report to solve the 20 analytical problems posed in this analytical project. The visualisations were grouped together based on their relevance to each other to provide an integrated dashboard outlook to help in improving the quality and reducing the speed of decision making (Saggi & Jain, 2018). This project seeks to support the organisational objectives of ABC by providing insight that can assist in targeting the audience, determining suitable time and day for the new videos to be uploaded and to upload the videos in the most suitable as well as visible categories of YouTube. It was identified from the findings that 161,470 videos were uploaded by 12,360 channels from four countries across 18 different YouTube categories. (24) Entertainment was the top most performing category whereas (43) shows were the least performing category. It was also found that the French channels are subjected to have less views as most of the viewers on YouTube are geographically located in the GB. Therefore, ABC should target viewers in GB across entertainment category.

9. Task 7: Reflection 9.1 Reflection by team member 1 I found the analytical project to be technologically challenging. I was not familiar with the user interface of IBM Watson Analytics due to which it took me few days to design the visualisations for first 10 questions. I learnt to use the user interface of IBM Watson Analytics and developed big data visualisation skills that can be used in future to complete analytical projects. I have also developed critical thinking and high level of collaboration skills as I was responsible to lead the other members. My contribution in this project was to distribute the tasks, download the dataset, uploading dataset to IBM Watson Analytics, to complete introduction, task 1 and visualisations for question 1 to 10 for task 2 reporting. 9.2 Reflection by team member 2 Designing the visualisation to answer the guided questions was the most challenging part of this analytical project from my perspectiv...