Reference Question Data Mining
Download as PPT, PDF1 like774 views
This document describes a methodology for mining reference transaction data to aid in reference, instruction, outreach, and collection development activities. The methodology involves librarians recording data from reference interviews, cleaning the data, classifying questions by subject, and collating the data into spreadsheets and reports. This process has led to the creation of subject guides, targeted library instruction, and informed collection development decisions at their university library.
Reference Question Data Mining
- 2. Traditional Reference Classification Directional (Where is the catalog?) Ready reference (How tall is Mt. Everest?) Specific-search question (Where can I find information about the progressive movement in Louisiana?) Research (lengthy detailed assistance)
- 3. Strengths Analyzes workflow Provides statistical comparison with peer organizations within the state Provides statistical comparison to evaluate national trends in reference services
- 4. Weakness Statistics reveal nothing about the content of the question This “reference data” is equally important as statistical compilation The reference literature contains no effective means of “mining” this data.
- 5. GOALS
- 6. Goals The creation of subject guides to address popular topics Target specific courses for library instruction Provide further data to aid in collection development
- 7. Methodology
- 8. Flow Diagram Record Data Clean Data Classify Data Collate Data
- 9. Recording Data After each reference interview the librarian records: Subject of question Class for which the information was sought Professor’s name Example: hair styles in the 1950s – ENG 102 / Costello
- 10. Recording Data It is the responsibility of each reference librarian to narrow the topic during the reference interview. Broad: “Decades” Narrow: “hairstyles” “1950s”
- 11. Recording Data Key words from the recorded questions come from the reference interview. Emphasis on recording the question and subsequent key words. Questions without recorded classes are usually identified as “community.”
- 12. Cleaning Data Sheets of reference data are collected monthly. Entries with incomplete, directional, or indiscernable data are excluded. Example: Directions to Starbucks Example: Business Week? Citations?
- 13. Classifying Data Entries are correlated to a call number range for each subject using Alice (Ohio University’s Library Catalog) and the Library of Congress Classification scheme Matching of relevant hits to call numbers is done at the discretion of librarian performing the classification
- 14. Classifying Data Data is correlated in ALICE (Ohio University’s Library Catalog)
- 15. Collating Data Subject headings and class numbers are compiled into an Excel spreadsheet. Professors names are not included in the compilation of the excel spreadsheet. Data is graphed visually for general overview.
- 16. Outcomes
- 17. Study Guides Reference Data Study Guides Reference Department
- 18. Study Guides First Subject Guide was “Utopia.” We now have nine subject guides directly related to our data mining and instruction initiatives. Agriculture, Decades, Environmental Science, Nursing, and so forth.
- 19. Outreach & Instruction Reference Data Instruction and Outreach Instruction Librarian
- 20. Outreach & Instruction Names of professors, courses, and related data are sent to the instruction librarian for outreach Example – Louisiana History / Professor Allured Almost one third of the student cap in her class are needing assistance at the Reference Desk.
- 21. Collection Development Reference Data Collection Development Collection Manager
- 22. Collection Development Monthly reports are sent to the Collection Manager Data is a component of the selecting process To date, the library has added several reference books and databases to the collection based on reference data mining
- 23. Reference Data Study Guides Collection Development Instruction and Outreach Reference Department Instruction Librarian Collection Manager
- 24. Biggest Obstacle to Implementation
- 25. Joshua Finnell Assistant Professor of Library Science Reference Librarian McNeese State University Walt Fontane Assistant Professor of Library Science Reference Librarian McNeese State University
Editor's Notes
- #15: This is a search in Alice for “hair styles.” Some of you familiar with the LC Classification may see US History (E 185), film, technology (TT972), and Geography (GT 2310). Since our root question was hairstyles in the 1950s, we are less concerned with technology and more concerned with Geography/Anthropology. To this end, GT 2310 is Geography/Anthropology/Recreation – Manners/Customs – Costume/Dress/Fasion.