Representation & Surrogation

LIS 5043: Organization of Information

Dr. Manika Lamba

Introduction

Two central concepts in the organization of information are representation and surrogation
Representation is a complex process. It is both the process and the product, depending on the context in which you are engaging with the concept
Surrogation is related to representation but can be thought of more as a process with a surrogate being the result of the surrogation process
Surrogates are objects that stand in place of something else, for example, a library card catalog record (MARC record) stands in place of the actual item in the library’s collection (such as book, cd, dvd) etc.
Whenever we create a surrogate, we have to decide what to represent (or to highlight) about the object and in the same instance, what NOT to represent about the object

Two central concepts in the organization of information are representation and surrogation. In this lecture we will explore both within a library context first, and also in contexts outside of a library.

Representation, as hinted about in the Overview lecture, is a very complex concept, and one that you will read about a great deal in the literature of the organization of information.

Surrogation is related to representation but can be thought of more as a process with a surrogate being the result of the surrogation process.

Surrogates are objects that stand in place of something else, for example, a library card catalog record (MARC record) stands in place of the actual item in the library’s collection, a book, cd, dvd, etc.

Whenever we create a surrogate, we have to decide what to represent (or to highlight) about the object and in the same instance, what NOT to represent about the object.

Representation is both the process and the product, depending on the context in which you are engaging with the concept.

Let’s take a closer look at the different forms of representation and how they relate to information organization.

Representation

Many types and levels of representation:
- Representation of user's information needs
- Representations of information objects
- Representations of relationships among objects
- Representations manipulated by systems
Ultimately, representations interact to connect users to information

The main forms of representation within the organization of information include:

Representation of a user’s information needs – whenever a user comes to an information system with an information need in mind, or a question to answer, or just a curiosity, they have to somehow represent that need within the system’s parameters. If they are using a library OPAC, they will have to know what fields are available to search. They will have to know how to enter the terms of their query into the system (what form of words to use, if truncation is possible, what punctuation to use, etc.). The search terms (query) that they enter into the system is a representation of their information need.
Representation of an information object – when users interact with an information system, they are, for the most part, NOT going to go directly to the information object itself as a result of a search, they will instead, find a representation of the information object (the webpage describing a product, the bibliographic record in a library catalog, the person’s human resources record, etc.) As noted on the slide, representations highlight certain aspects of an information object that we have decided are important or that are representative of the object.

For example, on an e-commerce website for a coffee shop, the person who created the site, may decide that not only are name, quantity, and price important to represent the coffee shop, but including the farmer stories who plant those coffee beans in the particular part of world is also important.

When we design information systems, we have to make choices about what to represent and what not to represent about the item.

These choices are informed by what we know about the objects, the users of our collections/products, and in general how users look for and use information objects

Representation manipulated by systems - records in a database, webpages on an e-commerce site, digital library records are all representations that are manipulated by systems.

As stated in item 2, when we design systems, we are informed by users and collections. We decide what fields to include in the database and records, what to display to our users on the pages, etc. We will talk more about this in the modules on users and systems.

Ultimately, representations are created in order to interact to connect users with information surrogates and with the objects themselves.

Model of Information Retrieval

Here is a model to help you think through the idea of multiple representations and how they interact to connect users to information.

Representation A is the user’s information need, which the user enters into a system

Representation B is the representation created as a surrogate for the information object

Both are entered into a system.

The retrieval component of the information system matches these two representations and bring back a subset of representations that the system contains that are representative of the words in the user queries.

We will not go into the mechanics of searching at this point, or how the retrieval mechanisms work within various systems. We will leave that for the module on systems.

The user then chooses a representation to review to see if it meets their information need or not.

Library Paradigm for Representation

Representation answers two questions

What is the item?
- Represent the container
- Bibliographic representation
- Descriptive, not evaluative
- Uniquely identify/describe an object
- Distinguish one object from another through metadata

What is the item about?
- Represent the content
- Subject representation
- Identify topics, concepts, perspectives, etc.
- Adequately represent object’s content so user can find the object based on search topic
- CONTEXT??

Other questions…How good is it? How will it be useful to me?

Two Primary Challenges

How to individually represent information objects as concrete entities so they can be found when needed?
- Container-oriented
- Uniqueness-oriented
How to represent and relate information objects as sources of information on various subjects so they can be found by those in search of the information?
- Content-oriented
- Relationship-oriented

Relations in the Bibliographic Universe

In libraries, we refer to objects that we can actually represent as being part of the bibliographic universe, an abstract subset of what is called the “Universe of Knowledge” in LIS.

Objects in the bibliographic universe are those that have been RECORDED in some manner/form. It does not include objects that are in a purely oral form (having not been recorded in some medium yet), such as spoken stories.

Once these stories are recorded onto audio tapes/video/digital forms, then we can organize them. While the term bibliographic universe stems from a library context, it can be extended to any context, as can the relationships between the objects in one’s collection. For example, in a corporation, the products produced by a company are what would be organized, say, company reports, white papers, working papers, concept briefings, emails of specific committees, etc. While we don’t think of these objects in the traditional sense of “collection”, they are indeed objects in a bibliographic universe that can be organized.

What other examples of contexts, collections, objects can you think of?

What other relationships?

Identifying Relationships

Separate containers, same content
- Different formats
Separate containers, similar content
- Translation, abridged, versions, editions
Seperate containers, same author, different content
- Link together all objects by author
Seperate containers, different authors, similar content
- Link together all objects by subject
Unique representations of container
Relationships established through various means
- Name authority control
- Controlled vocabulary

Bibliographic Tools (Instruments)

Mechanisms by which bibliographic control is established and maintained
Examples of tools (contain representations)
- Library catalog
- Bibliographic
- Indexes to journal literature
- Finding aids
Three functions of all bibliographic tools
- Identification or finding
- Collocation or gathering
- Evaluation or selecting
Representations must support these functions

In a library context, we call the creation of surrogates by several names, bibliographic control (control of the bibliographic universe), cataloging, indexing, etc. I personally like Patrick Wilson’s vision, “bibliographic harmony”—where we strive to provide a harmonizing effect between the representations in our catalogs/information systems so that users can find what they are looking for.

In libraries, we call the tools that we have created to effect bibliographic control (harmony), bibliographic tools or instruments.

There are many examples and I have listed a few here. The readings for this module will cover these in more depth and give you additional examples.

The functions of all bibliographic tools are to provide for: 1) Identification or finding 2) Collocation or gathering 3) Evaluation or selecting

These functions are the original objectives of a library catalog that Charles Ami Cutter developed back in the early 1900s. And they really have not changed much since. Abbas and Taylor will take you through some of the rich history of the development of bibliographic tools in the readings for this module.

Representations must support these three functions, in other words, they should include elements that help users identify or find a specific item in the collection or a specific subject; they should collocate or gather together all items by a specific author, title, or about a specific subject; and they should assist users in the evaluation or selecting of the object they need to satisfy their info need.

WHAT other functions can you think of that could be included in a surrogate in a library context? In other context? WOULD they really be any different?

Representation Responsive to Users

What to users want to do?
- Search?
- Evaluate?
How do our representations support their tasks?
Library catalog representations support four generic user tasks (FRBR)
- Find
- Identify
- Select
- Acquire or obtain

As noted earlier, representations should be responsive to your users. In order to design “user-centered” systems and representations, we have to know a little about how users interact with our systems, how they search, what they want to do with our systems and with representations.

Think about a general library catalog record. How do these representations support user tasks? A landmark study called the FRBR (Functional Requirements for Bibliographic Records) study was one of the first and one of the very few studies we have conducted to learn more about how users view our bibliographic records and information systems, as well as the tasks they hope to accomplish while using library catalogs/catalog records. Maxwell’s reading in a later module will talk more about FRBR. Also be sure to explore the FRBR and RDA websites to learn more.

You can see on the slide that three of the original objectives proposed by Cutter years ago still hold. FRBR studies identified an additional one: to acquire or obtain the object.

So, what needs to be present in a surrogate so that users can satisfy the FRBR tasks? Should the representation be more generic or specialized? Can you think of instances in which representations would be customized for specific user groups?

Metadata and Metadata Schemes

Data about an entity

Information in the library catalog

Metadata indicates something beyond the data: a statement or statements about the data

All metadata bears the imprint of its origins and intended use when examined closely enough

It is “structured information that describes, explain, locates, or otherwise makes it easier to retrieve, use, or manage an information resource” (NISO, 2017:1)
Metadata scheme consists of metadata elements
- Library catalog metadata scheme (MARC) contains metadata elements for title, author, subject, publisher, etc
Metadata element represents an attribute of the object

OK, we cannot leave this lecture without at least introducing the concepts metadata and metadata schemes. I know, I have been using these terms throughout this lecture but it is logical to present them here again, as the result of the data modeling we just did (identifying entity level and attributes) is the development of a metadata scheme. AND the data we enter in the record to describe the object is metadata.

The concept metadata is most often defined simply as “data about data” or “information about information”.

Librarians have been creating metadata for hundreds of years, we just did not have this sexy name, we called it cataloging.

I prefer to think of metadata as “structured data representing an information object”. We don’t simply add a few words to describe an object, we make informed decisions and follow specific rules and standards for how to create the representation or the metadata.

The term “metadata” was used before the advent of the Internet, primarily in computer science to specify information about database objects and/or program objects.

When you structure metadata into a system, you are actually creating a metadata scheme, or the field structure in a database, or the list of attributes or elements you will use to represent the objects in your collection. The MARC scheme we use in libraries is an example of an early metadata scheme. There are many more.

Purpose and Functions of Metadata

Purpose: represent an entity
Functions
- Discovering or finding that an entity exists
- Evaluating or selecting an entity
- Locating or accessing an entity
- Others such as collection management, rating content, managing intellectual property rights
Application areas
- Libraries
- Websites
- Government agencies
- Museums

The main purpose of metadata is of course to represent an information object.

However, because the term was coined for use in the digital environment (though it is also used in an analog context as well), the functions we saw earlier of a library catalog are slightly augmented.

A simple web search from most unsophisticated smart phone generate lengthy record of new data including

initial search term
location of your phone
time & date when you searched
terms you searched before/after
your phone operating system
your phone’s IP address
information of apps on your phone

We find present in some metadata schemes what is called “administrative metadata” or information that represents administrative activities associated with the object, such as when it was donated and by whom, who owns copyright and how might the object legally be used, any content ratings, what collection it is part of, etc. You might also find metadata specific to the digital capture of the object, such as scanner used, resolution of image(s), date scanned and by whom, etc.

Human Construct

Metadata is a human construct and not found in nature
It is designed to solve particular problem
Nothing objective about it
It makes subjective statement
- what should be included
- what should be omitted
- where to draw boundaries
- what terms can be used to describe it

Metadata does not exist naturally in the world. It is created by people to solve specific problems. Because of this, metadata is not objective. Every metadata system reflects human decisions about what information should be included, what should be left out, where boundaries are drawn, and which terms are used to describe things. In other words, metadata always makes subjective statements about data, even when it appears neutral or technical.

To understand this more clearly, think about metadata as data about an entity. A familiar example is a library catalog. The catalog doesn’t contain the book itself. It contains information about the book, such as its title, author, subject, and publisher. This information tells us something beyond the data itself; it helps us understand, evaluate, and find the resource.

Metadata always reflects its origins and intended use. If you look closely enough at any metadata system, you can see the assumptions and goals of the people who created it. This is why metadata created for one purpose may not work well in another context.

Metadata is organized using metadata schemes, which are structured sets of metadata elements. For example, a library catalog uses the MARC metadata scheme, which includes elements like title, author, subject, and publisher. Each metadata element represents a specific attribute of the object being described.

We are Data

We are filled with data in today’s networked society

through our web activity, we are assigned gender, ethnicity, class, age, education level, and potential status of parent with x no. of children (digital trace data/digital footprint/digital breadcrumbs)
if internet metadata identifies a user as foreigner than they lose right to privacy afforded to U.S. citizens
who would have thought that class status, citizenship, ethnicity could be algorithmically understood?

This book on “We are Data” by John Cheney-Lippold is a very interesting read which helps us to rethink what data and especially metadata means in our everyday lives.

In this networked society, we’re not just producing content; we’re constantly generating metadata through our web activity. That includes things like time stamps, locations, device information, search histories, clicks, and patterns of interaction.

From this metadata, online systems infer attributes about us, such as gender, ethnicity, class, age, education level, or even whether we are likely to be a parent. These inferences are not things we explicitly declare; they are derived from what’s often called our digital trace, digital footprint, or digital breadcrumbs.

What’s important to emphasize here is that metadata is not neutral. When internet metadata identifies someone as a “foreigner,” that classification can have real consequences. In the U.S. context, for example, non-citizens do not receive the same privacy protections as citizens, which means metadata can become a basis for differential treatment and surveillance.

This leads to the question: who would have imagined that class status, citizenship, or ethnicity could be understood algorithmically?

Metadata enables these classifications to be automated, scaled, and embedded into systems—often without transparency or consent. This is why studying metadata is not just a technical issue, but also a social, political, and ethical one.

Representative Statements

How should we create our statements?
We want statements that represent the aspects of information objects that will help users meet their goals
- i.e. to find, manage, select, locate, obtain
In a consistent way
- so that users can compare descriptions

Now, look at what makes a good representative statement. When we create statements, like metadata descriptions or records, we’re not just describing an object for its own sake. We’re doing it to support what users are actually trying to do.

Those goals usually include things like finding information, managing it, selecting among options, locating the right item, or obtaining access to it. So the question to keep in mind is: Which aspects of the information object will best help users accomplish these tasks?

Equally important is consistency. If statements are created in different ways, users can’t easily compare descriptions across objects. Consistent statements allow users to scan, evaluate, and make decisions more effectively.

Ackoff’s Pyramid helps us think about the difference between raw data, information, knowledge, and understanding. Representative statements are not just raw data – they’re structured, meaningful information designed to support use and decision-making.

Administrative Metadata

Metadata used in managing and administering collections and information resources

Examples
- Acquisition information
- Rights information
- Location information

Let’s go through different types of Metadata starting with Administrative Metadata.

Administrative metadata refers to metadata used to manage and administer collections and information resources, rather than to describe their intellectual content. It helps organizations keep track of resources and ensures they are handled properly throughout their lifecycle.

For example, acquisition information records how and when a resource was obtained, who acquired it, and under what conditions. This is especially important for budgeting, collection development, and accountability.

Rights information is another key type of administrative metadata. It documents who owns the resource, how it can be used, and any legal or licensing restrictions. Without clear rights metadata, organizations risk misuse or legal violations.

Finally, location information tells us where a resource is stored or how it can be accessed—whether that’s a physical shelf in a library, a server location, or a digital repository.

Administrative metadata helps us to ensures that information resources are organized, controlled, legally compliant, and accessible, making it a critical part of effective data stewardship and information management.

Descriptive Metadata

Metadata used to identify and describe collections and related information resources

Examples
- Cataloging records
- Finding aids

The second type of metadata is descriptive metadata. It is used to identify and describe collections and information resources so that users can find and understand them. Descriptive metadata answers basic questions such as what is this resource, who created it, and what is it about?

Common examples of descriptive metadata include cataloging records, such as those found in library catalogs, which provide information like title, author, subject, and publication details. Another example is finding aids, which are used to describe archival collections and help users understand the scope, structure, and contents of those collections.

The main purpose of descriptive metadata is discovery. It helps users search for resources, recognize their relevance, and distinguish one item from another. Without descriptive metadata, even valuable collections can remain hidden or difficult to use.

Preservation Metadata

Metadata related to the preservation management of collections and information resources

Examples
- documentation of physical condition
- documentation of preservation actions

The third type of metadata is preservation metadata. It is used to support the long-term management and preservation of collections and information resources. Preservation metadata focuses on ensuring that materials remain usable, authentic, and accessible over time, even as technologies and formats change.

Examples of preservation metadata include documentation of physical condition, which records the current state of an item, such as wear, damage, or degradation and helps guide conservation decisions. Another example is documentation of preservation actions, which records what has been done to preserve the resource, such as repairs, digitization, format migration, or environmental controls.

Preservation metadata is essential for maintaining the integrity and authenticity of collections and ensuring that future users can trust and understand the preserved materials.

Technical Metadata

Metadata related to how a system functions or metadata behaves

Examples
- Hardware and software documentation
- Technical details of digitization process
- Authentication and encryption information

The fourth kind of metadata is technical metadata. It is used to describe the technical characteristics of a resource and the systems used to create, manage, or deliver it. Technical metadata helps systems function correctly and ensures that digital objects can be rendered, processed, and preserved over time.

Examples include hardware and software documentation, which records the systems, platforms, and versions required to access or run a resource. Another example is technical details of the digitization process, such as resolution, file formats, compression methods, or scanning equipment used, information that is crucial for quality control and future preservation decisions.

Technical metadata also includes authentication and encryption information, which supports security, access control, and data integrity. Without technical metadata, digital resources may become unusable as systems evolve, making this type of metadata essential for reliable access and long-term management.

Use Metadata

Metadata related to the level and type of use of collections and information resources

Examples
- Circulation records
- Search logs
- User tracking

Finally, we have use metadata which captures information about how collections and information resources are used. Rather than describing what a resource is, use metadata tells us how often, in what ways, and by whom a resource is accessed or interacted with.

Examples include circulation records, which show how often an item is checked out or accessed, helping institutions understand demand and make collection management decisions. Search logs record what users search for, revealing patterns in information needs and helping improve discovery systems. User tracking data can show how people navigate systems or interact with resources, though it also raises important privacy and ethical considerations.

Use metadata is valuable for improving services, allocating resources, and evaluating impact, but it must be handled carefully to balance usefulness with user privacy and ethical data practices.

User-Generated Metadata

Metadata generated by users online

Examples
- Social-Bookmarks
- Reviews on Amazons and Goodreads
- Tags on LibraryThing
- Folksonomy (collaborative tagging, social indexing, social tagging)
- Crowdsourcing

https://crowd.loc.gov/get-started/how-to-tag

User-generated metadata is the metadata that is generated by users online, rather than by professionals such as librarians, archivists, or system designers. People can create metadata by interacting with digital platforms by tagging, reviewing, bookmarking, or otherwise describing content in their own words.

Some common examples include social bookmarks, customer reviews on platforms like Amazon or Goodreads, and tags on sites like LibraryThing. These activities are often described using the term folksonomy, which refers to collaborative or social tagging. Unlike controlled vocabularies, folksonomies are informal, flexible, and evolve organically based on how users think and talk about content.

Another important form of user-generated metadata is crowdsourcing, where large groups of users contribute descriptive information at scale. The image on the right shows an example from the Library of Congress’s crowdsourcing projects, where volunteers help tag and describe historical materials.

User-generated metadata can greatly enhance discovery and engagement by reflecting real user language and perspectives. However, it can also be inconsistent or subjective. This contrast helps us see how metadata creation involves trade-offs between control and openness, and why different metadata approaches serve different purposes.

Power of Metadata: Structure

Metadata Elements and Semantics

Introduction
Introduction (Cont.)
Dublin Core
Many Metadata Schemes

Each element needs a name
Each element needs a definition (semantics)
The semantics describe the information content associated with an element
Input rules associated with an element

15 elements to represent digital resources
Also used for analog resources

Government Information Locator Service (GILS)
- Metadata scheme for describing government information
- 67 elements
Federal Geographic Data Committee (FGDC)
- Metadata scheme for describing geospatial information
- 100+ elements
Encoded Archival Description (EAD)
- Used within the archival community
And many more….
- Visual Resources Core, MODS, etc.

Focus So Far, Upcoming Weeks

Structuring the representation
- Selecting object attributes to represent (based on users, user questions, and collection’s objects)
- Labeling attributes through the metadata elements/field names
- Defining the elements (semantics) to indicate information about object to record
In a few weeks we will try some data modeling