Final Assignment

 

Helen Kresl

 

 

 

 

 

 

 

 

 

 

 

 

 

IMT 530B - Organization of Information Resources

Winter Quarter, 2009

Mike Doane, Instructor

Table of Contents

Introduction. 3

Recap of Concepts. 3

Approach. 4

Implementation. 5

My Taxonomy. 5

Discussion. 5

My  Thesaurus. 6

Discussion. 7

Conclusion. 8

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Introduction

 

The first half of the quarter we learned about object based classification which is the application of metadata to information objects. We practiced formally identifying objects using available schemas ie. Dublin Core. We also began early stage controlled vocabulary building.  This involved choosing preferred terms and synonyms for our information objects with the goal of leading people to those information objects.  The second half of the quarter entailed more complex ways of classifying the information objects, this time not just to bring users to the objects but to shape their interactions with them. This final exercise involved ordering the information objects and then assigning hierarchical data to them for the purpose of controlling their “findability”.  The tools I employed for this were: Thesauri and Controlled Vocabularies (CVs).

 

Recap of Concepts

 

While in practice the words taxonomy and thesaurus are often used interchangeably, according to Stewart, taxonomy is just one step on the continuum towards building a complex CV and refers to just one level of complexity of classifying those objects.  At the bottom of the continuum is the synonym ring, which is just a list of words any of which can mean the same information object.  One step up, authority control identifies a single term within a synonym ring it as the preferred term for the object. Next up on the continuum are taxonomies which show hierarchical relationships with parent and child terms.  Thesauri, which are at the end of the continuum, use the approaches of all the previous methods plus that of using related terms ie. broader, narrower, and entry terms, for the most referential precision. This information is reiterated in the reading by Rosenfield and Morville (2002). See the below graph:

(Stewart pg 51)

Approach

 

My initial choice of domain (from the midterm) had to change, as it was advised that there were not enough information objects or relationships to work with on Language Services Associates.  So, I chose another domain called Territorial Seed, http://www.territorialseed.com, a website rich with many categories of information objects pertaining to the work and interests of a gardener.

The way I began thinking about the new set of information objects for this CV was essentially the same as before and based on Stewart’s advice to reconcile three often opposing forces, Content, Community, and Context. Content, is the collection of information objects on the website needing to be named; Community, is the user group for which the taxonomy is being built, and Context, refers to the constraints, such as budget, technology, or other resources that limits the scope of the CV. 

Content

Vegetable and Flower seeds:  This is a subset of all the content on Territorial Seeds.

Community

I thought about the community in terms of personas.  For example:  A customer coming to this website could be a beginner with an interest in growing vegetables in a few containers on her indoor/outdoor balcony receiving only partial sunlight, an experienced urban gardener with a primary interest in growing organic food for the kitchen, a person wishing to incorporate more flowers into her vegetable garden for beauty or for introducing beneficial relationships between flora and fauna for the purposes of pest control, or a landscape gardener intent on creating a dazzling color coordinated backyard for the city’s upcoming annual backyard garden tour.  The site has the raw materials for each type of gardener and it seemed that there was potential for organizing them in a way that would enable each of these types of gardeners to achieve their goals.

Context

My limitations of experience and time are probably the greatest constraints to this project.  The real purpose of it is for me to get practice with the steps involved in organizing information using practices and methods developed by others.

The ultimate goal was to create an order for the information objects that fit my conceptual model of what users would want and need when interacting with these objects. This meant deciding on which terms would represent the conceptual model.  In this way a “typical user” would find more than a repository of objects but objects ordered into a conceptual model of my choosing and in a controlled fashion. 

Implementation

 

My Taxonomy

 

http://students.washington.edu/hxkresl/imt530/Territorial%20Seed_taxonomy.xlsx

 

Discussion

 

“An effective taxonomy sits between Chaos and Order and mediates the two: it does not, as it’s so often assumed, represent the domain of Order unequivocally.” (Lambe pg259)

 

 

 

 

 


For the final I put more emphasis on drawing out the traditional characteristics of the information objects that might help a gardener when designing a contemporary garden. For example, with all the emphasis today on gardening as a hobby and ‘going natural’ I thought it helpful to be able to search against terms such as organic and bloom color in addition to types of vegetables. Also, by identifying the pest problems of certain plants the gardeners may readily find plants which attract insects that behave as natural predators of those plants prone to pest issues (hence ‘pest problems’ category) and grow them together.   By identifying bloom color I enable gardeners to combine flowers of colors of their choosing, and by designating a top category ‘months to harvest’ I enable the gardener to stagger crops for continuous harvesting. ‘The art of taxonomy is in discovering the client’s individual methods of interacting with information and combining this with the materials’ unique characteristics.  There are other possibilities that I didn’t draw out such as leaf patterns and sizes that would also be useful to a landscape designer, but the basic idea is established. Almost every line in the taxonomy represents a decision based on the needs of the client and the requirements of the information’ (Bertolucci pg 42)

I found that it was much easier to come up with groups of concepts for subsets of information objects before combining them into one spreadsheet, the ‘All Up’.  The end result was that top level categories for some concepts didn’t make to all the information objects.  You will see many blank lines in the ‘All Up Taxonomy’ that were not present in the subsets of CVs.  The only terms that mapped to all information objects were the administrative ones: GUIDs, Dated Added to Inventory, Expiration Date, and Quantity.

            In the end my taxonomy was shaping up to look like a classic poly-hierarchical taxonomy. EG:

1.      Vegetable or herb ‘is a type of’ flower.

2.       Vegetable ‘requires’ low, medium, high amounts of watering

3.       Plant ‘is an instance of’ heirloom, organic, or beneficial insect plant

My  Thesaurus

 

http://students.washington.edu/hxkresl/imt530/Territorial%20Seed%20CV.xlsx

 

 

Discussion

“To be worthy of the name ‘thesaurus,’ a controlled vocabulary must incorporate a minimum of three types of relationships, namely equivalence, hierarchical, and associative.” (Clarke 38)

 

 

 

 

 

 


The development of my CVs involved the use of:

1.      Scope notes:

2.       Entry or equivalent terms (ET) eg. misspellings

3.      Related or associative terms (RT) eg. organic = natural

4.      Broader terms (BT) eg. vegetable

5.      Narrower terms (NT) eg. latin terms

 

Each term restricts the definition of the preferred term so that it guides the users along a path to a destination of my determination, at the same time it also disambiguates the concepts that are being used.  Rather than having the word Amaranth lead a user to a floral shop, my thesaurus will hedge them in between ‘seed’ and the latin term for the flower.  Therefore, the user will not end up in a floral shop buying Amaranth but on our domain looking for seeds.  This is also known as user warrant. 

When considering the list of terms use as top level categories I was reminded to make them possess ‘equivalence’ which is the degree to which they represent equivalent concepts.  My top level terms: crafting and cutting, color of bloom, heirloom all refer to qualities that are considered when selecting plants to grow.  So even though “organic” has nothing to do with “color of bloom’ they are good top level categories.

            Semantic factoring posed the greatest challenge for me.  I spent much time considering what would be best suited for the ET, RT, BT, NTs.  I found I was using the same BT and NT for Things to Buy, Flower Type, Vegetable Type, and Herb Type which would meant they were all the same things which they aren’t so I refined. I did not find a narrower term for individual colors, or BT and NT for the words yes/no in the Organic thesaurus.

Conclusion

 

Our lectures introduced us to the work of other taxonomists and it was plain that there were so many ways to do them.   A helpful taxonomy for me came from Stewart, on page 164, where cheese is divided on the basis of softness.  While soft and hard are themselves not equivalent terms, in the context of describing cheese, they are, in that cheeses come in various densities.  This was a clear exhortation to find categories that made sense not by definition but by context.   The taxonomy most influential to my work was probably the ontology slide from lecture 7, slide 3. (source: http://ontolog.cim3.net ).   It was an ontology comprised of descriptive and administrative taxonomies that showed the beginnings of the how the relationships between concepts could become exploited and it was the first time that I began to understand the purpose behind taxonomy.

The goal now will be to practice the classification of ideas and objects and the existing tools for them so that I can take my idea, such as for Territorial Seed, and implement it in the real world.  Since the introduction of Ontology it has become clear to me that I do not need to understand the machines and machine languages that deliver the intellectual work of the field of ontology, nonetheless, I am still interested in exploring them.

 

 

 

 

 

 

 

Works Cited

Bertolucci, K. "Happiness is taxonomy: Four structures for Snoopy." Information Outlook, 7(3) (2003): 36-44.

Clarke, Stella G. Dextre. "Relationships in the Organization of Knowledge." Green, C. A. Bean & R. Thesaural Relationships. Oxon: C. A. Bean & R. Green, 2001. 37-52.

Gladwell, Malcolm. "Annals of Public Policy: Troublemakers; What pit bulls can teach us about profiling." The New Yorker (Feb. 6, 2006).

Lambe, Patrick. "The Future of Taxonomy Work." Organising Knowledge: Taxonomies, Knowledge and Organisational Effectiveness. Oxford: Chandos Publishing, 2007. Chapter 10.

Rosenfeld, L. & P. Morville. "Thesauri, Controlled Vocabularies, and Metadata." Information Architecture for the World Wide Web. Sebastopol: O'Reilly, 2002. Chapter 9 (p. 176-208).

Steward, Darin L. Building Enterprise Taxonomies. United States of America: Mokita Press, 2008.

 

Bibliography

Bertolucci, K. "Happiness is taxonomy: Four structures for Snoopy." Information Outlook, 7(3) (2003): 36-44.

Clarke, Stella G. Dextre. "Relationships in the Organization of Knowledge." Green, C. A. Bean & R. Thesaural Relationships. Oxon: C. A. Bean & R. Green, 2001. 37-52.

Gladwell, Malcolm. "Annals of Public Policy: Troublemakers; What pit bulls can teach us about profiling." The New Yorker (Feb. 6, 2006).

Lambe, Patrick. "The Future of Taxonomy Work." Organising Knowledge: Taxonomies, Knowledge and Organisational Effectiveness. Oxford: Chandos Publishing, 2007. Chapter 10.

Rosenfeld, L. & P. Morville. "Thesauri, Controlled Vocabularies, and Metadata." Information Architecture for the World Wide Web. Sebastopol: O'Reilly, 2002. Chapter 9 (p. 176-208).

Stewart, Darin L. Building Enterprise Taxonomies. United States of America: Mokita Press, 2008.

 

 

Instructor comments:

I really liked the way you stepped up and turned this project around from a disappointing midterm. You are well on your way now to being able to do this kind of work on a regular basis (if you wanted to, that is) with more confidence and clarity. Your report this time is clear and concise, and although I wish you would have better incorporated the readings within your writing, rather than pasted in as a separate thought, it still reads well and makes sense. Your CVs are a good start, although there are some typos and equivalency problems that I think you could have worked on more before submitting this. Certainly your ETs and RTs could have used some more work. I didn't get a real sense of how you wanted to fill them out, based on what I see there. Overall, though, a big improvement and I am glad that you stayed with the work and the class, even though at times I know it was tough. That takes a lot of courage and hard work, and I can see that in this final.   Final Assignment: 3.6  Quarter Grade: 3.7