Matthew Reidsma

Work Notes

Updates from the GVSU Libraries’ Web Team.
Archive // Subscribe: EmailRSS

What I did on my summer vacation: the creation of the GVSU Libraries knowledge base

This summer we were lucky to have Kathy Kosinski, a student at the University of Michigan’s iSchool, intern with us to work on the creation of a new knowledge base/self-help system for our users. Kathy spent 9 weeks combing through our data, talking with librarians and staff, writing copius amounts of documentation, and even playing a practical joke on me. It was great to work with her, and she’ll be missed. When she graduates in 2016, any library would be lucky to hire her.

Below is the final report I asked her to put together about the creation of the knowledge base. While the knowledge base isn’t yet live to our users (it will be in a few weeks, once we switch to the new version of the campus Content Management System) but you can preview it at Kristin is still working with staff at all campuses to ensure that we capture all the appropriate details for each campus, and we’ll constantly be iterating on making the knowledge base better. This has been a project that Kristin and I have talk about for a few years, and it wouldn’t have gotten off the ground without the drive and careful attention to detail that Kathy brought to the project.

And now, here’s Kathy: -M.

by Kathy Kosinski

With Kristin Meyer and Matthew Reidsma, I was tasked to come up with the content for a self-service help center for GVSU libraries. The process included data gathering and analysis, creation of governance documents, and, of course, the creation of the content itself. The following is a quick summation of the steps I took along the way.

Data analysis

Data scope

This data set includes the Grand Valley State University Libraries’ desk transactions, instant message transcripts, email interactions, and reference interactions for the period of April 15, 2014 to May 05, 2015. This time includes three major exam periods. The data was collected in order to guide decision-making for some of the more common questions asked in order for inclusion in a new knowledge base. Previously-identified “common questions” (e.g. printing, course reserve assistance, study room reservations, etc.) were set aside for immediate inclusion in the knowledge base. The data was then sorted through in order to determine the previously missed, high frequency questions. The next step in the process is the creation of governance documents that, together with this data, will determine the most effective presentational format for the knowledge base.

Combined data sets

Type of Question # of times referenced
Scanning questions 203
Item request (various) 176
Holds (various) 165
Community services 160
Fines on account/overdue notices 159
Where is “x” campus location? 140
General check out and return questions 137
Renewal requests 126
Alumni library access 122
Copy assistance 115
Multipurpose room 115
Poster printing 108
Library website comprehension questions 105
Writing center 89
Faxing availability 82
Document delivery/ILL 79
Knowledge Market/research consultant 76
Laptop rental 74
UBS/textbook queries 73
Location/reservation of computer lab 65
MeL questions 64
Hiring 58
Thesis/master’s projects formatting, binding, and where to find in the catalog questions 52
Chemistry Success Center 51
Materials available for checkout 48
Where are the librarians/where is a specific librarian? 44
Computer locations 43
Chargers 40
Blocked IP 36
Quiet space query/noise complaints 35
Argo Tea questions 32
Off-campus book delivery/return 31
Returning materials at a different GVSU location 30
iPad rental 28
Meaning of peer-reviewed/scholarly OR “Research Skills Tutorial” 28
General children’s books, location 26
Laminating 26
Tourist questions 25
Headphones 24
Atrium location 23
Library account access questions 22
3D printing/technology showcase questions 21
Off-campus access issues 20
Student Services questions 19
E-book access issues 18
Binding 16
Vending machine location 15
Microsoft Office 365 download 15
Exhibition room location 14
Purchase requests 13
Donation/gift policy 13
Bus questions 12
“Where are the books?” 12
Microfilm questions 12
Parking 12
Popular reading 12
Software available on library computers 12
Recall questions 11
Mailbox, campus location 10
Event boards questions 9
Spanish children’s books, location 9
Lego room/innovation zone 8
Scantron station questions 7
Curriculum Materials Library technology 7
Office of Undergraduate Research and Scholarship questions 7
Lighting issues 7
Law/Cooley questions 6
Microwave 6
Salary book 6
Public phone location 5
ATM location 4
Notary 4
Computer availability 4
GVSU Drive off-campus access 4
Food delivery questions 3
Streaming video issues 3
Inquiry to film in library 3
Heating 3
Mission statement 2
Tutoring Center 1
List of software available from IT 1

Desk interactions

Type of Question # of times referenced
Scanning questions 195
Where is “x” campus location? 138
Holds (various) 134
Community services 130
General checkout and return questions 117
Copy assistance 114
Multipurpose room 109
Poster printing 88
Writing center 84
Faxing availability 78
Renewal requests 74
Location of computer lab 62
Item request (various) 60
Fines on account/overdue notices 59
Knowledge Market/research consultant 58
Laptop rental 57
University Bookstore (UBS)/textbook queries 55
Alumni library access 51
Chemistry Success Center 51
MeL questions 43
Computer locations 43
Chargers 40
Where are the librarians/where is a specific librarian? 39
Library website comprehension questions 39
Document delivery/ILL 39
Materials available for checkout 37
Hiring 34
Thesis/master’s projects formatting, binding, and where to find in catalog questions 28
Argo Tea questions 28
General children’s books, location 25
Laminating 24
Tourist questions 24
Headphones 24
iPad rental 23
Atrium location 22
Returning materials at a different GVSU location 21
3D printing/technology showcase questions 19
Student Services questions 15
Binding 14
Vending machine location 14
Exhibition room location 13
Bus questions 12
“Where are the books?” 12
Microfilm questions 12
Mailbox campus location 10
Parking 10
Microsoft Office 365 download 10
Event boards questions 9
Quiet space query/noise complaints 8
Scantron station questions 7
Curriculum Materials Library technology 7
Office of Undergraduate Research and Scholarship questions 7
Popular reading 7
Spanish children’s books, location 6
Recall questions 6
Lego room/innovation zone 6
Donation/gift policy 6
Law/Cooley questions 6
Public phone location 5
Microwave 5
Off-campus book delivery/return 4
ATM location 4
Notary 4
Food delivery questions 3
Salary book 2


Out of 9,235 desk transactions, 2,491 fell into the category of unidentified, high frequency questions. The top questions fell into two camps: technology instruction and quick answer. Technology instruction (how and where to scan, copy, or fax) were the first, sixth, and tenth most frequently asked questions (respectively), and were routinely answered in the form of hands-on instruction from the desk worker. “Quick answer” questions all are characterized by having a short, finite answer:


  • How long will my books be on hold for?
  • How can I check if they are in?
  • Can I transfer my hold to another library?
  • Where can I pick them up?
  • What if someone else has it on hold?

Community Services

  • I am not a student here, can I use your computers/wifi?
  • I am not a student here, can I study here?
  • I live nearby, can I check out one of your books?

General Borrowing Questions

  • How many books can I check out?
  • How long can I check them out for?
  • Do I need my student ID to check them out?
  • Where do I check them out?
  • Where do I return them?

Scattered throughout the remaining questions are other campus services that overlap in the patron’s mind with library services. University Bookstore and textbook queries were fielded by the library desk, especially at the start of each semester. The IT Help Desk in the ground floor of the library offers materials for checkout (iPads, laptops, chargers, etc.) that differ from the materials that the library desk offers (books, dry erase markers, model skeletons, etc) and there is occasional confusion as to which desk to approach. Additionally the library service desk receives requests for materials available at other campus locations outside of the library such as media equipment from the Communications department or student transcripts from the Student Services building.

Chat transcripts

Type of Question # of times referenced
Item request (various) 63
Alumni library access 42
Renewal requests 40
Library website comprehension questions 40
Document delivery/ILL 32
Blocked IP (all queries) 27
Quiet space query/noise complaints 23
Community services 19
Thesis/master’s projects formatting, binding, and where to find in the catalog questions 19
Fines on account/overdue notices 18
General checkout and return questions 17
Off-campus access issues 17
Meaning of peer-reviewed/scholarly and “Research skills tutorial” queries 16
Library account access questions 16
UBS/textbook queries 14
Laptop rental 13
Poster printing 13
Knowledge Market/research consultant 12
Software available on library computers 12
Off-campus book delivery/return 12
Holds (various) 11
How to choose the correct database (on campus) 10
MeL questions 10
Hiring 10
Materials available for checkout 9
Returning materials at a different GVSU location 9
Lighting issues 7
Scanning questions 7
Microsoft Office 365 download 5
Writing Center 5
Student Services questions 4
Computer availability 4
GVSU drive off-campus access 4
Faxing availability 4
Donation/gift policyq 4
Popular reading 4
iPad rental 4
Argo Tea questions 4
Spanish children’s books, location 3
Heating 3
Parking 2
Salary book access 2
Mission statement 2
Purchase requests 2
Binding 2
Lego room/innovation zone 2
Where is “x” campus location? 2
3D printing/technology showcase questions 1
Tourist questions 1
Laminating 1
“How crowded is the library?” 1
Lab reservation 1
List of software available from IT 1
Multipurpose room reservation 1
Microwave 1
Atrium location 1
Exhibition room location 1
Copy assistance 1
Vending machine location 1


A total of 612 online chat interactions fell into the scope of the data collection. The most frequently asked questions in chat all related to digital interactions with the library. For example, requesting materials was the most common question. Patrons would check in to find the status of their requests, ask the library staff to request materials for them, determine why a specific item is “unavailable”, etc. These chat conversations would often start from within the catalog, with the chat acting as the next step for the patron to get their item. Similarly, the next five questions (blocked access to databases, confusion over website jargon or operations, issues with document delivery, and email notifications of materials’ due dates) all occur within the space of the user’s computer, and thus the patrons turn to chat in order to answer their resulting questions rather than via phone or in-person.

Email transcripts

Type of Question # of times referenced
Fines on account/overdue notices 82
Alumni library access 25
Item request (various) 23
Holds (various) 20
Hiring 14
Off-campus book delivery/return 13
Renewal requests 12
Purchase requests 11
Community Services 10
Knowledge Market/research consultant 6
Library account access questions 6
Recall questions 5
Poster printing 5
Subject specialist/liaison librarian 5
Blocked IP (no classes/unknown) 5
Multipurpose room reservation 5
E-book access issues 5
Laptop rental 4
Quiet space query/noise complaints 4
Document delivery/ILL 4
IP blocked (current student, faculty, or staff) 4
UBS/textbook queries 4
Streaming video issues 3
Library website comprehension questions 3
Queries to film in library 3
Donation/gift policy 3
Materials available for checkout 2
Thesis/master’s projects formatting, binding, and where to find in catalog questions 2
Meaning of peer-reviewed/scholarly 2
Book checkout, general 2
Lab reservation 2
Children’s books, location 1
Book return, general 1
Salary book access 1
“Research skills tutorial” query 1
3D printing/technology showcase questions 1
Laminating 1
MeL questions 1
Scanning questions 1
Tutoring center 1
iPad rental 1
Popular reading 1


The email interactions were characterized by a significant drop in frequency between the most frequently asked topic (with 82 occurrences) and the second (with 25 occurrences). The data shows that users overwhelmingly turn to email to answer their questions in regards to fines and overdue materials. In fact, email communication about fines was higher than the other contact methods combined (a total of 77 occurrences).

The other popular email queries fell into several categories including digitally-centered questions (alumni library access, item requests, renewal requests, recall questions), distance-affected questions (alumni library access, off-campus book delivery/return, purchase requests), and queries with email attachments (hiring).

Reference transactions

Type of Question # of times referenced
Known item searches 30
Library website comprehension questions 13
MeL and ILL search instruction 10
Meaning of peer-reviewed/scholarly 7
ILL rejection help 4
Alumni library access 4
Thesis/master’s projects formatting, binding, and where to find in the catalog questions 3
Off-campus access issues 3
Poster printing 2
Salary book access 1
Community services 1
“Research skills tutorial” query 1
Off-campus book delivery/return 1


Out of close to 1,200 reference interactions within the date period, only about 80 fell into the data collection criteria. This is useful as it demonstrates the liaison librarians aren’t getting bogged down with the mundane, “quick answer” type questions on a day to day basis.

Yet they did experience some quick, ready reference as the highest category in this data set was “known item searches.” These interactions featured the librarians using their own personal Summon-workaround knowledge, or online searching experience to find an article that the patron already had the title and other information for.

Creation of governance documents

In order to help shape both the creation and ongoing maintenance of the knowledge base, we created a plan for content strategy. The document covered the following criteria:

  • Inclusion of content
    • initial creation of the KB
    • subsequent additions
  • Maintenance
    • new data analyses
    • review of current KB content -criteria for update/deletion
    • protocol for updates
    • protocol for deletion
  • Criteria for success
  • Voice and tone guidelines
  • Style decisions
    • formatting for titles and links
    • preferred phrasing

With these guidelines, we are able to keep up a (hopefully) sustainable plan for the knowledge base: something that would allow the content within to stay relevant, correct, and manageable.

Initial inclusion

Ultimately, we chose to include knowledge base topics that incurred at least ten user queries in a one year period over multiple service formats (e.g. desk, chat, email, reference). By these standards, our final number of entries for the knowledge base totaled 52.


The list of knowledge base entries was split between two GVSU librarians for “ownership” after my departure. Per the governance documents, the entire knowledge base would be reviewed annually by staff at each library location. The staff would flag problematic entries as “Update Needed,” and the librarian in charge of that entry would begin the updating process.

Voice and tone

A major goal for the knowledge base was to get the voice and tone of speaking down correctly. We used the MailChimp voice and tone guide for inspiration, but we also found some helpful advice from Courtney Greene McDonald and her book Putting the User First. In the end, we decided on a few guidelines:

  • Be concise, convenient, and straightforward in the presentation of information.
  • Be informal without being disrespectful. Speak in “business casual”
  • In most cases, use personal pronouns like “you” and “we”
  • Highlight the library’s goals of inspiration, flexibility, and support
  • Keep language positive, even if you are saying no.

Content creation

Categorization of data

We aimed to have the content accessible both by browsing and searching. To facilitate browsing, we had to categorize the data in such a way that there weren’t an overwhelming number of categories when the visitor first landed on the knowledge base home page.

Using other knowledge bases available as a comparison point, we again looked to MailChimp as a jumping off point. I then began categorizing the data with a general goal of creating a handful of useful, viable categories.

Using an affinity wall style of categorization, I started by putting similar user needs and questions together in groups. From there, I looked for larger, more general groups. Rather than four categories like MailChimp, we ended up with three:

  • Need some help?
    • This covers questions about where to get academic help, such as research or tutoring assistance
  • Using the library
    • This section details some of the more logistical questions we receive at GVSU, such as parking, hours, study rooms reservations, etc.
  • How to find
    • Here users can find out how to request items, use the library search, and more.

The content journey…

This is depressingly one of the shortest portions of the final summary. Depressing because ultimately it was the most time-consuming portion of the knowledge base creation. Namely because good content is hard; even good content that follows a premeditated content strategy.

The process of creating the content involved searching both GVSU websites and speaking with faculty and staff about the answers to many of the user questions and problems encountered. My process involved finding the answers, writing a response, and then returning later to review the draft. Each entry of the knowledge base required several revisions and voice and tone checks.

Each entry went through several iterations of editing and were proofread by both Kristin Meyer and Matthew Reidsma. The next step in the content phase will be carried on after my departure, it includes having other GVSU librarians outside of the knowledge base creation proofreading the entries that relate to their specialties.

Tagging the content

Each entry had to be associated with a handful of tags that would allow the content to be searchable in the knowledge base. Initially the tags were guided by “what felt right” in terms of the language I had seen used over the course of all the interaction data I had gone through. I furhter tailored these by looking at the searches performed through Summon, the library’s e-resource search.

The tags received another user-based boost from the results of a series of cognitive mapping tests run by Kristin Meyer and other GVSU librarians. Using her results, I was able to incorporate some user’s wording and descriptors for different library locations. The exhibition space, for example, was never referred to as such, but rather was spoken in terms like “the big glass art room.” I was able to add to the knowledge base entry on reserving library space and include terms which may not have seemed applicable at first glance.

Shortcomings of my work

Due to the time constraints of my internship, I was unable to test the content on library users. Given more time and higher library traffic, I would have set up informal testing in the entry areas of the Mary Idema Pew and Steelcase libraries in order to test the content in the system with actual users. This test would have involved several versions of a handful of entries, each version written in different voice and tone. The users would then pick which entry they found to be the best.

One option I have suggested for after my departure is the inclusion of an additional form along with the already in-place procedures for recording the results of patron interactions. I suggest that the form (sample) act as a way to track whether the content in the knowledge base is effectively answering the sorts of questions library users are asking. Grading the content in this way is important, as there is really no point in funneling users towards a knowledge base if we aren’t sure that we will be able to sufficiently answer their questions once they get there.