INF385T/CS395T: Topics in Information Retrieval and Web Search (Fall 2010)
The University of Texas at Austin

INF385T/ CS395T (cross-listed): Topics in Information Retrieval and Web Search (Fall 2010)

Instructor: Matt Lease
Day and Time: Fridays 1-4pm
Location: UTA 1.212 in the iSchool

Previous offering: Spring 2010

Course schedule and readings


In an Information Age promising instant access to seemingly limitless digital information, search has become a ubiquitous paradigm for enabling information access. However, creating an effective search engine requires meeting a variety of important practical challenges:

  • Characterizing the nature of search relevance (both topical and user-oriented)
  • Defining operational models for ranking based on relevance
  • Developing search algorithms which are accurate, scalable, and efficient
  • Designing interfaces, interaction mechanisms, and graphical visualizations providing an engaging and user-friendly search experience (human-computer interaction and visualization)
  • Understanding trade-offs between system-oriented and user-oriented methods for search evaluation
Information Retrieval (IR) studies both human information needs and the systems built to meet those needs. As such, IR has lain squarely at the intersection of Information Science and Computer Science since its inception. IR studies methods for capturing, representing, storing, organizing, and retrieving unstructured or loosely structured digital information, as well as designing interface, interaction, and visualization methods for creating an effective and compelling search experience. While digital information was once restricted to electronic documents, today's landscape of digital content is incredibly rich and diverse, including Web pages, news articles, books, transcribed speech, email, blogs (and micro-blogs), images, and video. The rise of the Web as a massive, global repository and distribution network has earned Web search engines and other Web technologies particular importance in organizing and finding information today.

The course will culminate in a one-day end-of-semester Workshop at which students present their course projects to showcase their work.

Course Objectives

  • Provide broad exposure to the field of IR via weekly readings and in-class discussion of IR topics (breadth)
  • Develop expertise and first-hand experience in a particular specialized topic of IR via a course project conducted individually or in small groups (depth)
  • Develop professional skills essential for both research and non-research activities:
    • Efficient and critical reading of published scientific literature
    • Making effective presentations: in-depth and elevator pitch
    • Verbally communicating scientific knowledge across disciplines (and potentially to the public at-large)
    • Performing online literature search for prior scholarly work
    • Organizing and managing a project
    • Scientific writing for scholarly publication
Student Responsibilities
  • Read several published IR papers each week and write a short (8-10 sentence) critique of each to concisely (1) summarize key ideas, (2) identify main contributions, and (3) discuss strengths/weaknesses
  • Once or twice a semester, present an assigned reading to the class and moderate discussion
  • Develop a major course project, individually or in small groups, with the goal of publishing the work. The instructor will advise the work and expect students to engage in independent exploration of concepts and execution of tasks. Meta-ideas:
    • Term paper: write a survey on state-of-the-art practice in a specialized area of IR (review and synthesis of published scientific literature)
    • Algorithm: implement and evaluate a new search algorithm
    • Analysis: Present a novel analysis of one or more existing IR systems
    • User-centered evaluation: evaluate IR system effectiveness via user-oriented qualitative and/or quantitative methods (e.g. interactive IR, task-completion accuracy and/or times, usability issues, affective perceptions, etc.)
    • Human-computer interaction: design a new search interface, implement/mock-up, and evaluate
    • Visualization: design a new graphical visualization method for conveying search results or managing information overload
    • Crowdsourced evaluation: explore crowdsourcing methods for informing or evaluating search engines
    • Mobile IR: Develop a mobile IR application using our pool of Google Android phones

No prior knowledge of IR or programming expertise is required; all interested and motivated students are invited to attend. This course typically attracts significant student participation across a wide variety of disciplines: information science, computer science, linguistics, electrical engineering, and design studies. Course activities are intended to serve the needs of both (1) those studying to work professionally on search engines or conduct research in IR, and (2) non-specialists interested in gaining broader exposure and understanding of IR methods and systems.

Undergraduate seniors may enroll only by permission of the instructor.

Course Textbook: none required, all readings available online


  • First day information form

Online Books

New version of Baeza-Yates books is forthcoming

Other References

Related courses at UT Austin

External IR information and resource pages