Presenting Big Data with Interactive Data Visualization Tool
Välimäki, Vesa-Ville (2017)
Metropolia Ammattikorkeakoulu
2017
All rights reserved
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2017092015263
https://urn.fi/URN:NBN:fi:amk-2017092015263
Tiivistelmä
Analyzing big data is not easy with lightweight tools. Recent data leaks have shown that exploring data with traditional methods takes too much time. This thesis investigated whether today's data visualization JavaScript libraries can create agile and interactive user inter-faces. Browsers are facing hard times when massive amounts of data are analyzed on the screen.
The study was done in three parts during the year. The history of data visualization was studied first and after that the most well-known cases in the industry were investigated. Web technologies that have boosted the development were also under the research.
In the second phase, the best JavaScript libraries supporting interactive data visualization were selected. The examples provided by the libraries were studied in terms of speed, user friendliness and future development.
At the end of the studies, the best JavaScript libraries were heavily pushed to handling 100,000 elements on the screen. The survey material used was massive and open data provided by the San Francisco City and the County. This huge data file contained property in-formation from San Francisco City area.
The study focused on the interactive side of the issue. This means how many nodes and edges visualization can take so that the user could still analyze the data smoothly. Interactivity was the key element of the study, so the drawing capability of static nodes and edges was not measured. The number of edges used in the tests was held in 500 pieces at all times. The study also marked out how the performance of the computers or the different browsers affected the results.
The final results were in line with the technical specifications JavaScript libraries promised. Today’s JavaScript libraries are able to produce, together with the modern Internet browsers, a sufficient number of elements to the screen without slowing down. Best libraries are also able to offer good editing opportunities to users who lack sufficient computing skills.
The study was done in three parts during the year. The history of data visualization was studied first and after that the most well-known cases in the industry were investigated. Web technologies that have boosted the development were also under the research.
In the second phase, the best JavaScript libraries supporting interactive data visualization were selected. The examples provided by the libraries were studied in terms of speed, user friendliness and future development.
At the end of the studies, the best JavaScript libraries were heavily pushed to handling 100,000 elements on the screen. The survey material used was massive and open data provided by the San Francisco City and the County. This huge data file contained property in-formation from San Francisco City area.
The study focused on the interactive side of the issue. This means how many nodes and edges visualization can take so that the user could still analyze the data smoothly. Interactivity was the key element of the study, so the drawing capability of static nodes and edges was not measured. The number of edges used in the tests was held in 500 pieces at all times. The study also marked out how the performance of the computers or the different browsers affected the results.
The final results were in line with the technical specifications JavaScript libraries promised. Today’s JavaScript libraries are able to produce, together with the modern Internet browsers, a sufficient number of elements to the screen without slowing down. Best libraries are also able to offer good editing opportunities to users who lack sufficient computing skills.