Results from power BI report

I explored how Power Automate can be used to combine the use of multiple Microsoft Services. My primary task was to see how data collected using a form could be stored, visualized, and shared.

Users filling in the form can submit their email. Since there are multiple forms, it was necessary to set up a system so that the same email address is not stored multiple times. Using Power Automate, it is quite easy to add data to an Excel file. In addition, it is possible to read all data in a table, and check if any existing emails match the new email. So, if there are no duplicates, the new email is added to the Excel sheet.

My next objective was to create a visualization of people’s answers to a question, then show it with anyone who filled in the form. For this, I first created a chart in Power BI using the data collected from the form. Howver, I found that publishing reports on the internet is a Pro feature. As I am a free user, I cannot share reports online.

I opted to use Excel again for the visualization. I found a very helpful tutorial by DamoBird365, linked here, which shows how scripts can be used to automatically generate charts based on existing data. I tested this by creating a workflow to send myself emails whenever anyone filled in the form.

This week, I created a web scraper to collect auction data from multiple auction websites. I used python libraries like BeautifulSoup, results, pandas and cvs to do web scraping and visualize the collected data.

I began by learning about the basics of web scraping and common ethical considerations related to the field. Then I tried following some tutorials to scrape data from websites made for learning web scraping. I particularly liked this tutorial by Real Python, which introduced concepts in an engaging manner and helped me understand the rationale behind each step.

For my project, I collected information about the auctions of works by a single artist only. I collected this data from the websites of Bonhams and Sotheby’s.

I quickly learned that the methods I used to collect html snippets before were ineffective for collecting information from websites that use javascript to load front-end content. I mainly worked using the BeautifulSoup library, which is insufficient for scraping javascript-rendered pages. So, I used Selenium, a software which automates web applications. It offers another advantage - while many webpages load content only when a user scrolls, Selenium automatically scrolls to the end. Therefore, when data is scraped using Selenium, the entire HTML file is loaded.

In Bonhams’ website, all information about auctions related to a specific artist can be searched for. At first, I collected data from this page only. However, the information available on this page is quite limited - only the title of the work, the link of the auction, as well as the estimated price can be found. I wanted to learn when the paintings were auctioned, so I then visited the url given for each work, and collected additional information. Using the information obtained, I created tables and downloaded them as Excel files. I used a similar methodology to collect information from Sotheby’s’ website.

You can see some screenshots of my work below:

The work I have done is far from perfect, and I think I could imporve results by dedicating more time to this project. However, I am happy I was able to learn a new skill, and look forward to using this knowledge in the future.

I explored open access collections from museums around the world - the Rijksmuseum, the Metropolitan Museum of Art, the M+, and more. Many museums have made their vast collections of public domain artwork free for use, allowing anyone with an internet connection to view and utilize high-resolution images and metadata. People can use these services for education, research, commercial purposes, or for entertainment. The New York Public Library published an article online demonstrating how people can use works from their collection, including making GIFs! (Lampasone)¹

I was excited about working with the open access collections of different museums. I began by exploring what projects and services people have already created using them. I loved some of the games found on the Google Arts and Culture website, like the Art Coloring Book, where users can color famous pieces of art as they like, and Pet Portrait, which compares photographs of one’s pet to that of an animal in a painting! I also loved seeing then-and-now comparisons by different museums, which show how different a location looked in the past. You can check out some here!

I worked on creating a website that could show the timeline of an artist’s works. It uses the open access APIs of The Metropolitan Museum of Art and the Cleveland Museum of Art (CMA). The APIs are used to collect information about the works made by a particular artist, and create a timeline.

The API of the Met and CMA have different structures, so I used two separate functions to query data. I began by working using the Met’s API, which conveniently has a function for searching for artists. However, I soon found that any artist whose name contains matching words are included in the returned result. For example, if I searched for ‘Vincent van Gogh’, I would get works from him along with other artists with ‘vincent’ or ‘van’ in their name. So, I added some conditions so that the results would be filtered correctly. This is an example of the working filter:

I then worked on fetching data from the Cleveland Museum of Art’s API in a similar way. After combining the two and experimenting a bit, I noticed that not all works display images. This is probably because some works are protected by copyright law. However, basic information about the works is still displayed, so I decided to keep them in the timeline.

There are a few limitations of the website. The main problem is that it does not work if I search using a keyword that does not match the value returned by the API precisely. For example, if I search for ‘Henry Edmond Cross’ instead of ‘Henry-Edmond Cross’, I do not get the expected results. This applies to accent marks too. Another drawback is that if there are a lot of pieces in the museum’s collection, it takes a fairly long time to generate the timeline - up to a minute or more!

You can view my website here 🌸

Here are some snippets of the timelines generated using my project for different artists:

Van Gogh:

Michelangelo:

Julia Margaret Cameron:

This week, I worked on creating a publication for the exhibition Tales of A Connected World using Quire. Quire is an open-source tool created by Getty that allows the same set of files to be used for creating a website, pdf, and epub.

You can view the pdf here.

Creating a new quire project automatically uses files from the quire starter default repository, which can be viewed on this page.

I loved using quire, because it simplifies adding different types of media, including image carousels, icons, audio, and video files. This is also the first time I have created an ebook, something I never thought I would do, so I felt really happy about working on this project.

Another advantage of working with Quire is that it supports multiple languages. I have tried adding Arabic text, and it worked without any issues. The only problem is that quire is built to support primarily English, so all text elements are aligned left-to-right. While plaintext content can be changed into right-to-left alignment easily, it is difficult to do so for metadata. I think that it would not be difficult to implement such a change, but it would be quite time-consuming and require some trial and error.

I started this project by reading through quire’s documentation and watching a video tutorial. However, I quickly ran into a problem - I got errors when I ran ==quire build== command, which is necessary to ensure that pdf and epub can be created. I visited Quire’s Community Forum and found that it is a common bug in Windows, and implemented the workaround suggested.

The generated pdf did not have any images, aside from the cover image and a few others. By default, paged.js is used to generate pdf. Quire’s docs suggested an alternative approach of using PrinceXML to generate the pdf, which I hoped would fix the issue. However, pdfs generated by both libraries did not display images.

To find a solution, I read through the relevant parts of Quire’s documentation as well as their forum, and tried different approaches to solve the problem. I found a potential answer - using html tags directly instead of shortcodes. But that displayed the images only on the pdf, not on the website. Moreover, I wanted to use quire’s built-in resources, like image carousels. If I switched to using html tags, I would have to implement everything myself, which would make using Quire pointless.

I have submitted a request for help on the forum, and received a prompt response, so I hope a solution can be found soon.

Some screenshots of the website are displayed below.

This week, I built an audio guide for Lusail Museum by using the audio guide template published by ACMI Labs. The repository for the original audio guide can be found here.

You can check out my website by scanning this QR code or clicking this link

When a visitor opens the website, they are greeted by a welcome page with simple instructions for using the audio guide. Then they are taken to the home page, which allows them to choose between 2 languages - English and Arabic, and choose which audio track to listen to. They are then taken to the page for the selected audio track, which contains the track and a transcript. They can switch to another track using the audio player, or the side menu.

My experience developing the audio guide

At first, I went through the documentation on ACMILabs’ Github repository, which explains the general setup and layout of components in a clear, concise manner. I do not have much experience with forking existing repositories and working on them, so perusing their documentation and going through their source code was very helpful. I also opened the live demo website, and compared it to the source code to familiarize myself with how different components are arranged, which was quite helpful later on for customizing the website’s appearance.

I extracted audio files from several videos using VLC Media Player for this project. I began by setting up the page for the first audio track. This included adding a title, transcript of the audio file and images.

I did some trial and error, but eventually, the page looked as it should. I then added some more audio track pages in a similar fashion.

The next feature I wanted to incorporate is giving users the choice to switch between English and Arabic. I did so by including buttons on the home page that takes users to the corresponding language’s homepage. The homepage then displays only the audio tracks in the user’s chosen language.

This solution is not perfect, but the other attempts I made at establishing this feature were not successful. At first, I tried to use javascript to switch between languages. The advantage of using javascript is that it would allow me to easily modify the site’s behaviour by simply changing the value of a variable. However, none of the approaches that I tried worked. So, I used a simpler solution - buttons that link to another page.

Finally, I tested some features to make sure everything was functioning properly. To make sharing the website easier, I generated QR codes for several pages. Later, I shared the website with several of my colleagues and got their feedback.

Some pictures of the completed website:

I truly enjoyed working on this project, and hope you like the audio guide!

On my first day, I met the people in the Lusail Museum team, and got to know more about their work by attending the weekly meeting. It was interesting to learn what everyone is working on and how they do different tasks, especially the preparation for future exhibitions. Later, Mr. Christophe showed me and Jenan many of the exhibits and the different types of multimedia content available in the National Museum. It was a truly wonderful experience, and it increased my interest in learning and contributing as much as I could during this internship.

Later, I observed how visitors interacted with multimedia in different galleries of the museum and noted down my observations. Many visitors seem to like the large projections on the walls, and it’s a popular subject for photographs.

On Monday and Tuesday, I talked to visitors about their experience in the museum. This was a novel experience for me since I have not conducted in-person interviews before. I was a bit apprehensive about it, but the process was surprisingly enjoyable. I got to meet people from different countries, and hearing the visitors’ thoughts on their visit, as well as how and why they chose exhibits/media to engage with was interesting too.

The results of the survey revealed some interesting information. Most visitors are tourists, and I expected most tourists to learn about the museum through social media or a tourism-related website. However, that is applicable for only half of the tourists among the participants; the other half said that they became interested in visiting the museum due to a personal recommendation from a friend, colleague or family that lived in Qatar. Overall, the participants rated the multimedia in the museum highly, but they interacted with different types of multimedia to different extents. The large projections on walls, wooden models and the noninteractive screens beside displays are the most popular, whereas the least popular form is the audio tour, which most people seem to not know of.

On Thursday, Mr. Martin gave me a tour and explained the behind-the-scenes work that goes on to make all the projections work seamlessly. He explained how the projectors are set up to account for the large and sometimes irregularly shaped walls, how video footage is edited and color corrected, as well as how everything is maintained. It was very informative, and I respect all the work that went into setting everything up, and the regular upkeep required.

My first week here has been really exciting and I am glad I got to try out so many new things! I look forward to what I will learn and work on in the coming weeks.