Qui Vivra Verra - Initial Visualizations & Findings
Features
The dashboard visualisation below allows users to understand the flow of patrons to each of the libraries using the FY2013 datasets. The following section will explain the different features in the dashboard.
- Location of Library
The map visualises the location of the selected library.
- Geographical Distribution of Patrons
The map visualises the flow of patrons (by Planning Area) to the selected library.
- Distribution of Patrons
The bar chart is sorted in descending order to identify the top Planning Areas with higher number of patrons for the selected library.
- Distribution of Patrons by Clusters
The distribution of patrons by clusters (from RFM Analysis below) for selected library is as shown.
- Number of Nearby Amenities From Library
The number of nearby amenities (MRT stations, Malls and Tuition Centres) within 1 km from the selected library.
- Number of Patrons in Library
The number of unique patrons visiting the selected library.
Findings
- Community Library vs Regional Library
Through the initial visualisations, the team has discovered some patterns in the Patron Dataset provided. The team has observed different patterns of distribution of patrons for community libraries and regional libraries, where majority of the patrons from the community libraries tend to come from only the Planning Area closest to the library, whereas majority of the patrons from the regional libraries are dispersed across more Planning Areas nearer to the library.
For example, comparing the patron distribution for Jurong West Library (community library) and Jurong Regional Library (regional library), 65.91% of the patrons at Jurong West Library are from Jurong West, where the Planning Area is closest to the library.
On the other hand, majority of the patrons to Jurong Regional Library are dispersed across more Planning Areas such as Jurong West, Jurong East and Bukit Batok which are located nearer to the library.
Regional library tends to have a larger collection of materials and a greater floor space as compared to the community library, which may help explaining why there are more patrons coming from different areas of Singapore. However, both community and regional library are usually more localised, thus drawing patrons who are living nearby.
- Libraries Located in the Central Region
Unlike the trend observed in the previous section where patrons tend to live near to the libraries, the team has discovered that libraries such as Central Public Library, library@chinatown, and library@esplanade that are located in the central region tended to draw in patrons from many different areas across Singapore.
The trend is apparent in all 3 libraries where the distributions of patrons are quite even distributed across Singapore and this could be due to the accessibility to the libraries and the amenities around the libraries.
All 3 libraries have a considerably high number of amenities as compared to the other libraries as shown below:
Features
The dashboard visualisation below allows users to understand the flow of patrons from each planning area using the FY2013 dataset. The following section will explain the different features in the dashboard.
- Location of Library
The map visualises the patron distribution across the country in terms of planning areas. Each polygon on the map is a planning area of the country.
- Geographical Distribution of Patrons
The map visualises the flow of patrons from a planning area to all libraries.
- Distribution of Patrons
The bar chart is sorted in descending order to identify the top Libraries with higher number of patrons from the selected region.
- Distribution of Patrons by Clusters
The distribution of patrons by clusters (from RFM Analysis below) for selected planning region.
Findings
- Proximity to Library
Through the initial visualisations, the team has discovered some patterns in the Patron Dataset provided. There are different patterns of distribution of patrons for each planning area, where the majority of the patrons for each planning area tends to go to libraries near to their planning area.
One example of this trend is the behaviour of patrons living in the Ang Mo Kio planning area. From the visualization, we can see that the majority (47.31%) of those living in Ang Mo Kio tended to go to the Ang Mo Kio public library.
This trend remains the same for both public libraries (as shown above) and regional libraries (as shown below).
Therefore, the proximity of libraries to the planning area may be a significant factor to explain the patronage level to the library from the planning area.
- Ease of Travel to a Library
Although we have established that proximity to a planning area is a key factor for determining patron traffic to a library, our visualization has uncovered another key factor, besides proximity, which is ease of travel from the planning area to the library. This trend is uncovered when the team attempted to visualize the patron flow for those living in Mandai planning area. The visualization is shown below.
As there are no libraries in the Mandai planning area, patrons living in that area would have to travel to one of the three libraries in the surrounding planning areas instead. These libraries are Yishun Public Library, Woodlands Regional Library, and Sembawang Public Library. As Mandai is located such that it is equidistance to each of the 3 libraries, we expected that traffic to all three libraries would be approximately the same. Contrary to expectations, we see that that Yishun Public Library had the largest portion of Mandai patrons at 50.45%, nearly five times that of Woodlands Regional Library (10.04%) or Sembawang Public Library (9.87%). This can be attributed to the fact that there is a direct bus route to Yishun through Mandai. Therefore, we can conclude that ease of travel to the library may be an important factor in explaining the difference in patronage levels between libraries.
Features
After conducting clustering on the patron data set using the three attributes, Recency, Frequency, and Monetary (denoted as R, F, and M respectively), we have obtained 6 clusters, which we have visualized using Tableau. The following sections will explain the different features in the dashboard.
- Patrons in Cluster
The numbers below the cluster number shows the number of patrons in that cluster.
- Characteristics of Cluster
The text below the number of patrons in the cluster shows the characteristics of that cluster.
In this case, it shows that this cluster is characterized by patrons who borrowed books recently, at a frequent rate, with a large number of books borrowed at one time.
- Distribution of Patrons in Cluster
The distribution of patrons based on each attribute of R, F and M.
When selecting a cluster, the corresponding region in the bar charts will be selected to show the position of the patrons in the cluster.
In this case, cluster 5 is selected. The highlighted portion shows that a large number of patrons in cluster 5 have R, F and M values that are above the median.
Findings
- Clusters for Patrons in Library
The cluster distribution for patrons in a library is generally similar to the one shown below for Ang Mo Kio Public Library.
We can see that there are more active patrons (Clusters 1, 3 and 5) than non-active patrons (2, 4, 6). Patrons in the active group are characterized by high Recency, as well as high Frequency or Monetary. They can be considered active because usually, if one borrows frequently, they would need to borrow less books at a time, and vice versa. The non-active groups are characterized by low Recency, or low values for both Frequency and Monetary. This trend persists for both regional and public libraries.
An interesting trend can be observed for the Chinatown library, in that the proportion of patrons in cluster 3 is significantly less than usual.
The same trend can be observed for Esplanade Library.
This shows that patrons going to these libraries tended to visit them very often.
- Clusters for Patrons in Planning Areas
The cluster distribution in planning areas are generally similar to those for libraries, in that clusters 1,2 and 3 are more prominent. This is shown below with Ang Mo Kio planning area as an example.
However, there are some interesting patterns that can be observed for the few planning areas that deviate from this trend.
We can see above that the Boon Lay planning area is dominated by patrons belonging to cluster 4, having lows values in R, F and M. These are patrons that have not borrowed a lot of books, and have not visited the library for a long time. More attention could be paid to these patrons in order to re-engage them.
