Scribd
Upload
34 views29 pages

Dr. Asif Naveed - PE For Library Professionals

Uploaded by

inam rehman
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
34 views29 pages

Dr. Asif Naveed - PE For Library Professionals

Uploaded by

inam rehman
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 29

Prompt Engineering for Library

Professionals

Dr. Muhammad Asif Naveed


Department of Information Management,
University of Sargodha
PhD, MPhil (Gold Medalists), MSc (Gold Medalist).
MS Public Policy (Gold Medalist)
Road Map
• Defining Prompt Engineering (PE)
• How does PE work?
• PE Evolution
• Prompt Engineering as a Discipline
• PE Applications in Libraries
• Prompting Techniques: Basics + Advanced
• Prompt Anatomy and Types
• Best Practices
• Risks and Challenges
• Future of Prompt Engineering
• Concluding Remarks
What is Prompt Engineering (PE)?
• A process of crafting an effective input
for AI systems to get desired output.
• The art and science of crafting effective
prompts to guide large language
models (LLMs) towards desired outputs.
• The process of structuring an
instruction that is interpretable and
understandable by a generative AI
model.
• Emergence of more advanced models
• Future of PE is bright as human-AI
assistant skill
• Revolution started in 2020 with an
introduction of GPT3
• Exponential power of prompts was clear
and visible with stunning outputs
• PE became must-have skill

• Google BERT laid groundwork for transfer


models
• Showed how pre-training can produce
more robust LLMs
• Allowed more complex development of
GPT engineering.

• Pioneering LLMs sparked the idea for


prompting to generate useful text
• GPT/PE were limited to trial & error
Prompt Engineering as a Discipline
• An Emerging Discipline
• Focuses on designing and
optimizing text-based prompts to
elicit specific responses from AI
systems.
• Goal is craft clear, concise, and
effective prompts to obtain relevant
and credible information form AI
models.
Key Aspects of Prompt Engineering
• Understanding AI Models: Capabilities, Strengths, Weakness,
Baises
• Linguistic Expertise: Strong grasp of language, including syntax,
semantics, and pragmatics, to craft clear and concise prompts.
• Task-specific knowledge: Require domain-specific knowledge to
design prompts that are relevant to a particular task or
application.
• Experimentation and testing: Must test and refine their prompts
to ensure they elicit the desired responses from AI models
• Critical thinking, creativity and problem-solving skills
• Ethics and fairness: Consider the ethical implications of their
prompts and strive to design prompts that are fair, unbiased, and
respectful.
• Updating
• Summarizing • Data insights
recent and • Article
articles, AVs • Creating • Personalized • Generate
relevant data writing
• Translations FAQs product reports
• Product or • Blog posts
• language • Customer recommenda • Data
weather • Create quizes
learning Support tions visualization
updates • Coding, etc.
• Concise info charts
Why is Prompt Engineering for Libraries?
• Enhanced Search and Discovery
• Personalized or Tailored Services.
• Streamlined Library Operations
• Innovative Library services
• Research Support Assistance
• Digital Collections (Repository, eBooks, etc.)
• Artificial Intelligence Literacy
PE for Information Retrieval
• Improving Search Queries
• Using natural language queries instead
of keywords.
• Incorporating user context and
preferences.
• Discovering Hidden Information
• Identifying and extracting valuable and
relevant information from digital
collections.
• Uncovering connections between
different sources
PE for Personalized Services
• Creating User Profiles
• Gathering information about user
interests and preferences.
• Using this information to tailor
recommendations and services (CAS, SDI)
• Developing Interactive Experiences
• Creating chatbots and virtual assistants
for personalized guidance
• Designing gamified learning experiences
PE for Streamlined Operations
• Automating Repetitive Tasks
• Generating catalog records, creating shelf labels,
and managing circulation.
• Improving Knowledge Management
• Organizing and retrieving information from
internal databases and e-collections
• Creating knowledge bases for staff training and
development
PE for Innovative Information Services
• Creating AI-Powered Tools and
Resources
• Developing chatbots for reference and
research assistance.
• Designing interactive storytelling and
educational experiences.
• Enhancing User Engagement
• Using AI to create personalized reading
lists and book recommendations
• Developing gamified learning experiences
Provide
Context & Examples
Background info

Specify output
format
Clear Task
Definition
Prompt Types
• Informational: used to retrieve or summarize data in research and academia
• Example: “Summarize the latest research on digital archives.”
• Creative: Helpful to craft an engaging content for outreach and communication
• Example: “Draft a promotional email for a new library event.”
• Procedural: ideal for creating checklists or guides for operational tasks
• Example: “List the steps to digitize a collection.”
• Analytical: Critical to gain insights and make data-driven decisions.
• Example: “Analyze borrowing trends for the past year.”
Basic Prompting Techniques
• Clear and Concise Instructions
• Use specific keywords and avoid ambiguity.
• Provide context and background information.
• Show, Don't Tell
• Give examples of desired outputs
• Use templates and formats
• Iterative Refinement
• Experiment with multiple prompts to
optimize results
• Learn from the LLM's responses
Advanced Prompting Techniques
• Few-Shot Learning
• Provide a few examples of desired
outputs within the prompt.
• Chain-of-Thought Prompting
• Guide the LLM through a step-by-step
reasoning process
• Role-Playing
• Assign roles to the LLM and the user
Zero-Shot Learning
One-Shot Learning
Few-Shot Learning
Chain-of-Thought Prompting
Best Practices for Prompt Engineering
• Start with Clear Goals
• Define the desired outcome of the prompt
• Be specific and Clear
• Balance simplicity and complexity
• Know Your LLM
• Understand the capabilities and limitations of the LLM you are using
• Experiment and Iterate
• Try different prompts and refine them based on the results
• Train and Evaluate
• Document Your Process
• Keep track of successful prompts and share them with others.
Risks and Challenges of Prompt Engineering
• Unpredictable outputs
• Over-reliance on AI applications
• Ethical and Legal Considerations
• Privacy and Security Concerns
• Transparency and accountability
• Technical Complexity
• Cost Considerations
Resources for Library Professionals
• Online Courses and Tutorials
• Learn the basics of prompt engineering
and advanced techniques.
• Community Forums and Groups
• Connect with other library professionals
interested in AI particularly in PE
• Vendor-Specific Documentation
• Access documentation and support for
specific LLMs and tools.
The Future of PE in Libraries
• Continued Innovation:
• New and emerging applications of
prompt engineering in libraries
• Collaboration and Partnerships
• Working with AI developers and
researchers to advance the field
• Ethical and Responsible Use
• Ensuring that AI is used ethically and
responsibly in libraries
Concluding Remarks
• Prompt engineering is a powerful tool
for library professionals
• By mastering prompt engineering,
libraries can enhance services, improve
operations, and create innovative
experiences for users
• Embrace the potential of AI, but use it
ethically and responsibly.
Thank You

You might also like