CDLI Search Results Visualizations
CDLI has rich geographical and temporal data at its disposal. Previously, this information was not fully utilized. Although the data schema is being improved, there were significant challenges in exploiting the new relationships available.
Through this project, temporal and geographical data are presented to users in an interactive manner, giving them a new way to browse and discover information. The temporal and geographical data are coupled with other information such as text genre, language, region in which they were discovered and periods, and displayed through D3.js, a JavaScript library for producing dynamic, interactive data visualizations in web browsers.
My major contributions during GSoC can be found in the following links
- https://gitlab.com/cdli/framework/merge_requests/21
- https://gitlab.com/cdli/framework/merge_requests/17
- https://gitlab.com/cdli/framework/merge_requests/16
Some other contributions can be found here.
- Create Pluggable visualizations for search results using D3.js
- Bar Chart
- Donut Chart
- Radar Chart
- Line Chart (with Brush Zoom and a
selectdropdown feature to toggle between different categories via AJAX update) - Dendrogram Chart
- Choropleth Map (with Zoom In/Out and Reset features)
- Integrate the visualizations to the advanced search results page
- Construct general stats page which gives the overall stats review of the entire database
- Develop Stats Controller for retrieval of required data from the database to be displayed
- Update UI of the framework
- Document the entire code written during GSoC 2019
- Add
noscriptfallbacks to be displayed when js is disabled
- Revamp UI of the framework
- Develop Bar Chart
- Add Zoom In/Out and Reset feature to the Choropleth Map
- Add Select dropdown feature to toggle between different categories in the Line Chart
In this GSoC project, 6 different visualizations along with a general stats page have been constructed, all of which are responsive to screen sizes. Appropriate fallback images together with a friendly text message had also been added to be displayed when a user has disabled JavaScript, which is a requirement to toggle all the visualizations since they are built on top of D3.js, a JavaScript Library.
Apart from that, few selections of attributes combination have also been specified as a select dropdown in the Line Chart and the Choropleth Map so that our users can view different combinations of attributes in the same visualization easily.
Besides that, an important aspect of this project is that the code is written in a very modular way so that it can easily be generalized to new attributes without many difficulties.
Although the queries to retrieve data from the database are optimized in the best way possible from my side, there is still a scope of little improvements by minimizing the time and space tradeoffs at a few places for one or two visualizations.
Further, the set of attributes used currently can only explain the information about certain categories of data selected to be used on the visualization via the voting done in this Google form. It would be better to try other remaining attributes available in our CDLI database too in order to provide more information to our users in a visual, easy and interactive way.
This GSoC project will help the CDLI users in getting a quick and clear understanding of the information about the artifacts related to their searches. The visualizations will help the user in analyzing large volumes of artifacts related data in an understandable and coherent way, which in turn helps them draw conclusions and insights without going through entire data which is almost impossible (as we currently have 334,540 artifacts and 1,538,442 publications). This will save the time of visitors to our site and help them identify emerging trends in the data. Without these graphical representations, relationships & patterns within digital assets and highly correlated parameters cannot be identified. Also, the human brain tends to process visual information far more easily than written information, providing faster comprehension of relationships and longer remembrance of data statistics.
The ability to statistically visualize a very large historical corpus will also aid CDLI in garnering further interest from academics and general audiences.
I plan to be a long-time contributor to CDLI and wish to mentor the organization in coming year events if given the opportunity. I would also be available in case any help is required with further improving my current project or in general with any task which falls under any of my areas of expertise.
I had an amazing experience during GSoC 2019 with CDLI. My mentors were very helpful and supportive. They guided me with whatever help I required and suggested various ways in which I could improve my work so that it can be helpful to a large number of CDLI users including those who are differently-abled.
Further, I also learned a lot about D3.js and most of the related things which can be done with the power of D3.js.
- Developer - Amaan Iqbal
- Mentors - Émilie Pagé-Perron, Rune Rattenborg, Shraddha Manchekar