Class 10 – Creating Metadata Rich Information Services

Part I Overview: Digital Libraries

The term ‘digital library’ is as ambiguous of a term as ‘metadata.’ There are a wide range of digital library and archival systems, each of which feature their own purposes, metadata standards, digital object management functions and domain specializations. In this course our goal is to understand information organization systems and their use in libraries, archives and museums. As part of this, this class includes an exploration of a digital library system whose purpose is to visualize metadata created and managed in other systems. This semester-long project involves metadata schema selection and design, metadata creation/harvesting, user needs consideration and interface design. In this worksheet we explore the digital library platform that we will be working with over the course of the semester and briefly introducing the elements of the assignment.

Instructions:

Work individually or in groups to complete the worksheet. When you get to a section that requires you to select a resource to explore – pick one resource (please don’t always choose the first one!). When asked to ‘discuss as a group’, consider your response and continue completing the worksheet.

We’re going to work with computer coding today and here’s an important note as you follow the exercises. Computer code is shown on numbered lines and are enclosed in boxes. The numbered lines are simply to help as a reference during instruction and should not be copied into your program. For example a line that reads 56. p { visibility:hidden; } should simply be typed in as p { visibility:hidden; }

Suggested readings

  1. Mitchell, E. (2015). Chapter 8 in Metadata Standards and Web Services in Libraries, Archives, and Museums. Libraries Unlimited. Santa Barbara, CA.
  2. Does Discovery still happen in the library?
  3. Watch: Making Stuff Faster. David Pogue. Available at
  4. Read: An online Magna Carta: Berners-Lee calls for bill of rights for web | Tech | The Guardian.
  5. OAI for beginners – the Open Archives forum online tutorial.
  6. In conjunction with the worksheet: review the viewshare tutorial:

Optional Readings

  1. Cole, T., Mischo, W., Ferrer, R., Habing, T., & Kraft, D. H. (2000). Using XML, XSLT, and CSS in a digital library.Knowledge Innovations: Celebrating Our Heritage, Designing Our Future. (Proceedings Of The 63Rd ASIS Annual Meeting, Volume 37), 430-439.
  2. Lagoze, Carl and Herbert Van de Sompel. (2001). The Open Archives Initiative: Building a low-barrier interoperability framework. Joint Conference on Digital Libraries..
  3. Woodley, Mary S. (2012). “Crosswalks, Metadata Harvesting, Federated Searching, Metasearching: Using Metadata to Connect Users and Information.” Introduction to Metadata.

Day 1: Register for a ViewShare account!

Step 1:Register for an account with ViewShare. Visit and Sign UP. Note, your account can take a day or two to get approved.

What is a digital library?

A digital library is a very large concept. It includes the elements of metadata storage, digital object storage, multiple user interfaces (e.g., Administrative, Discovery, Data-publishing). Digital libraries involve Controlled vocabularies, thesauri, taxonomies, classification schemes, metadata schemes and a host of Information Organization practices and technologies to manage these activities. In our exploration of ViewShare we are engaging with a pre-built digital library that supports a number of metadata management, discovery and metadata publishing tools.

Digital libraries serve a wide range of roles. In many cases, institutions implement a single digital library that contains all of their digital objects and metadata. In other cases, they implement multiple systems, each of which is best suited for a particular type of data. For example, a university may use a metadata system such as Archivists Toolkit to store their archival metadata, an institutional repository such as Dspace to store digital objects, a publishing platform like Omeka to present specialized collections of information and a discovery interface that supports resource discovery and use. In selecting the platform for a digital library organizations are faced with a wide range of decisions, each of which your group will consider this semester.

Creating a digital library

The creation of a digital library is a complex process that includes selection of resources, analysis of user needs, modeling of metadata, cataloging and interface design/implementation. These steps need not always occur in this order but in general it is useful to consider user needs before you create the metadata as well as the public interface! In this worksheet we explore the process of creating a digital library using ViewShare, a hosted digital library service provided by the Library of Congress. In this exercise we will skip past topic content/selection and user needs assessment but I encourage you to think about these steps as you complete the exercise.

  1. Key elements
  2. Topic/content selection
  3. User needs assessment
  4. Metadata model selection / definition
  5. Data loading / cataloging
  6. Interface design and implementation

Topic/content selection

Topic or content selection is often driven by external factors including what your LAM institution has available, what you want to make available to users or how you want to portray your collections. Key considerations in content selection include asking questions surrounding copyright (e.g. do you have the copyright or permissions to make this content available digitally), content fit with a digital library platform (e.g. Does your DL system support digital videos or just still photos) and questions of impact (e.g. is this the right collection to put online now).

User needs assessment

When selecting content or a DL focus it is also important to consider real-world use cases. This should include a definition of the community your digital library serves, the topical interest of the user and how it relates to the library content and the metadata and information interaction needs of the community. Having a clear vision of these uses at the beginning of a DL process will help shape your selections of metadata standards as well as drive the interface design at later stages.

Metadata model selection / definition

Your selection of a metadata model should fit the metadata needs of your community and relevant content in your digital library. At the same time, your standard should be selected so that it works well with the data model in your Digital Library platform which in this case is ViewShare. Although in theory our metadata needs should drive information system design, in reality this is often not the case given investment in an existing digital library platform or other technical or organizational constraining factors. ViewShare supports a number of data types including text, URLs, image, date/time, location and number. These data types can be implemented in any flat metadata structure.

Lets quickly familiarize ourselves with ViewShare by exploring an existing dataset.

Step 2:Go to Download the file and open it

ViewShare does not have a cataloging interface. It is designed to be a discovery and data visualization platform. In order to load data into ViewShare you create a spreadsheet or CSV file, an XML file using the MODS metadata schema or harvest objects through an OAI harvester of ContentDM connection.

In this worksheet we will work with creating metadata in an excel spreadsheet but you may also want to work with a MODS XML file for your project. When you upload a document to ViewShare it uses the metadata elements in the XML file or the first row in your excel/csv spreadsheet to create the field names. It also attempts to automatically detect the data types of your metadata (you can modify this later).

MODS Case study

MODS or Metadata Object Description Schema was created by the Library of Congress to represent bibliographic metadata (e.g. MARC metadata) but is useful for a wide range of metadata purposes and is similar to both MARC and Dublin Core. MODS is similar in some ways to Dublin Core in that it focuses on a limited set of core elements (20) and is designed to be a lossy but universal metadata platform. MODS is unlike DC in that it differentiates between some elements that DC does not. For example, while DC only refers to topicality through the <subject> element, MODS includes both <subject> and <classification> elements. Likewise, while DC uses the <type> element to refer to document type, MODS includes an element called <typeOfResource> which has an associated set vocabulary including text, sound recording-musical, still image and moving image (among others). MODS also includes the element <relatedItem> which can have any or all other MODS elements as children. This quasi-structural form of metadata is very different from DC’s flatter approach to description in which fields such as Identifier would be used to indicate these relationships.

The MODS standard is located at the Library of Congress web site ( If you want to work with MODS in creating metadata for ViewShare you link to the MODS schema at ( You could create XML records using eXchanger.

Spend a few moments reviewing the metadata in the excel spreadsheet you downloaded and answer the key questions below.

Key Questions

Question 1.What descriptive metadata fields are in the spreadsheet?

Question 2.What data might be useful for a map view? A timeline view?

Question 3.Does this metadata conform to a known schema? If you were going to apply a metadata schema (e.g. MODS, DC, QDC, EAD), which one would you choose?

Select schema and catalog records

In this exercise we will not catalog records but in your project, you will select a schema and create an excel file or MODS file that matches this schema. It is most likely easiest if you work in excel and perhaps even use the Fairfax data as your guide for creating data that ViewShare will easily map into its data types. Note, Instead of column headings like Name, City and State, etc., you should select metadata fields from a schema.

Exploration of metadata and interface design

Take a moment and explore a few published ViewShare sites:

These two collections make use of some of ViewShare’s data visualization features including Mapping and Timelines. In addition, they use information representation structures to take advantage of classification metadata in the database. Once you have reviewed the two collections answer the key questions below.

Key Questions

Question 4.What classification structures are used (e.g. taxonomies, Call Numbers, Faceted entries)?

Question 5.How are other data visualized (e.g. timelines, maps, charts)?

Question 6.With this in mind, how would you visualize the data in your spreadsheet?

Data loading and cataloging

A note on digital objects and ViewShare

ViewShare is an interesting example of a Digital Library in that it does not support storage of digital objects. It only allows you to load a URL that points to a digital object. This is a perfectly acceptable approach to Digital Library development but can pose issues in that in addition to ViewShare you must also have a place to store these images. Most libraries have a digital object storage system that serves this purpose but sites like flickr and picasa are also options!

Upload your data to ViewShare

Step 3:Lets start working with our sample data by uploading the example spreadsheet to ViewShare.

  1. Login to ViewShare (
  2. Click on the “Upload data” button
  3. Under “Simple Spreadsheets” select “From file on your computer”
  4. Select your Fairfax postcard xls file and click upload
  5. Take a moment to review the metadata

Key Questions

Question 7.How did the metadata from the spreadsheet get imported? Were there any errors?

Question 8.What data types are available under “type”? What is the importance of typing data in this way?

Tag and normalize your data

Step 4:Lets start by changing the metadata types for appropriate fields. Find the URL field for images and change that to “Image”. Likewise, change dates to date/time. Are there other fields you need to change?

Step 5:Lets turn some of the fields in our data into “normalized” or “system actionable” data.

  1. Click on the “Add” button, type a name for a new Location column and click on the Map icon
  2. Click the + next to City/State
  3. Click Create
  4. Back on the main data input screen click the augment button

Step 6:Repeat this process for timeline and tag cloud fields. Make your own determination about the fields that need to be selected

Step 7:Review the changes to your metadata

ViewShare uses a process called Data Augmentation to turn structured text metadata into system actionable data. For example, when you added a field that used the Map function and passed the City and State field to it, ViewShare looked up your location information in a database and returned Lat/Long information. Likewise, when you mapped the Image URL to an Image field, ViewShare knew to use code to retrieve and display the image rather than just a link to the image.

Key Questions

Question 9.What happened when you “Augmented” your data? Were there any errors? Why or why not?

Step 8:Save your data and get ready to create some “views.” In saving you will have to provide a name and description and decide if your data should be public. You do not need to make your data public but you will need to make your view public so the class can see it.

Interface design and implementation

Views are Viewshare’s public interface. Each dataset can have an unlimited number of views. We will experiment with creating one view that features many of the ViewShare elements

Step 9:From the data screen (the default screen once you save a file) click on Build

  1. Select a canvas for your view. For this exercise, choose 2 column left justified.
  2. Take a moment to look around the view design page. Experiment with adding widgets and views.

ViewShare includes a number of view development options grouped under widgets (elements that are designed to be contained in the sidebar navigation) and Views (data displays that take up the center of the page). Your final project should implement multiple views and widgets. Your digital library should contain all of the information required in the project assignment including a scope statement of your project, a brief statement of your user-needs focus and data views to showcase your metadata

Key questions

Question 10.What views can you create with ViewShare? What metadata in this collection is capable of working with those views?

Question 11.Imagine that you had three faceted subject fields in your metadata, one for location, one for date and one for topic. What views and widgets could you use to provide access to these?

Step 10:Lets create some specific views

  1. Click on Add Widget, select list and add a list for category and date.
  2. Click on Add Widget and add a Tag Cloud
  3. Click on Add View and add a Map View
  4. Under Map, examine the dropdown menu options. Experiment with them some (refer to the screenshot to the right for suggested values).
  5. Once you are done, look at views again. Add timeline and gallery views.
  6. Notice that you can move your views around (the left most view is the default!)

Key Questions

Question 12.What type of classification scheme does the list widget in ViewShare implement?

Question 13.What type of classification scheme is often used to represent tag clouds?

Data Visualization

Data visualization is the process by which users of information systems can study data by interacting with a visual representation of that data. Data visualization requires highly normalized and computer-usable data and uses computational techniques to produce interactive charts, graphs, timelines and other data-rich structures. As you explore ViewShare, you will see a number of different visualization tools.

Question 14.Spend a moment experimenting with these tools and match the type of data with the tool used to visualize it. Potential values for the metadata type column include “Categorical data”, “Categorical and Numeric data”, “Dates,” “Geographic information”

Table 1 Map of metadata types and visualization techniques

Visualization technique / Metadata type
Pie Chart
Line / Bar chart
Map
Timeline
Tag Clouds

Add contextualizing information to DL

So far we have focused on working with the metadata for our digital collection. It is equally important however to work with interface design and collection level data as well.

Step 11:Edit your view

  1. Under Widgets explore the text and logo options.
  2. You can use this URL to help you format text:
  3. Under Themes explore the different themes

Step 12:Once you are finished exploring these options publish your view.

  1. Click on save for view
  2. Click on Share, and then generate a link
  3. You will use this approach to turn a link to your digital project into your instructor.

Key Questions