Unit -6

Introduction

Information exploration should be enjoyable, but many users struggle with overload and
anxiety. Advances in digital libraries are helping users navigate vast information spaces more
easily, thanks to improved search and visualization tools.

As information grows, finding specific items can be challenging. Larger sources, like books or
libraries, can overwhelm users. Information specialists are developing strategies to help
narrow searches, making it easier for everyone to find what they need.

Multimedia documents add complexity, containing images, audio, and video. Digital libraries
provide curated collections, while archives may be less organized. Directories offer metadata
to guide users to specific resources. With various tasks from fact-finding to exploration,
flexible tools are essential to enhance the user experience.

Specific fact finding (known-item search)

Extended fact finding
Exploration of availability
Open-ended browsing and problem analysis

Searching in Textual Documents and Database Querying

The way people search online has transformed, allowing everyone from students to
researchers to find information easily. This accessibility marks a shift from expert-only
searches to a user-friendly experience.

Search engines like Google have improved through systems like PageRank, which assesses the
importance of web pages based on links. This leads to more relevant results while starting
with a simple interface.

However, the quality of retrieved documents still requires improvement. Users can follow
links to specialized information, but enhancing the relevance of search results remains a
priority.

Database searches are widely used for tasks such as booking travel, grocery shopping,
and accessing digital libraries.
Specialized databases help professionals like lawyers and scientists find relevant
information quickly.
tructured Query Language (SQL) is a standard for querying relational databases, enabling
specific searches based on attributes like author and publication date.
QL's power requires training, and users often make mistakes when writing queries.
 Natural-language queries aim to simplify searching but frequently lead to frustration due
to limited processing capabilities.
Form-fill-in queries simplify user input while still allowing some Boolean combinations.
Inconsistent search interfaces force users to relearn search methods each time they
switch systems.
Standardization of search interfaces could improve user experience, similar to how
automotive design evolved for safety and consistency.

The success of a search service hinges on minimizing user frustration. A five-phase framework
can enhance user experience for various levels of users accessing textual and multimedia
libraries:

1. **Formulation**: Provide access to relevant sources, allow entry of phrases and name
recognition, and enable variants to relax search constraints.
2. **Initiation of Action**: Use consistent buttons for explicit actions and enable implicit
actions that trigger new results based on parameter changes.
3. **Review of Results**: Offer explanatory messages, overview previews, and options to
adjust result sets, sequencing, and explore clustering.
4. **Refinement**: Guide users with meaningful messages for refining searches, make
parameter changes easy, and incorporate relevance feedback
5. **Use**: Allow users to save, annotate, and share queries and results, integrating them
with other tools for further analysis.

Multimedia Document Searches

Search interfaces for multimedia-document libraries are still developing, relying on
descriptive documents or metadata for locating images, videos, and sounds.

Without adequate captioning and human annotation, finding specific items, like a ribbon-
cutting ceremony photo, can be challenging, though automated filtering can help.

Future advancements must integrate powerful annotation tools, improved search algorithms,
and effective browsing techniques to enhance user experience in multimedia searches.

Image Search :locating specific images, like the Statue of Liberty, is difficult due to
variations in orientation and lighting. Query by image content (QBIC) targets distinctive
features, but it's easier with restricted collections. Effective tools like PhotoMesa
enhance personal photo browsing.
Map Search :Online maps allow searches by latitude and longitude, while geographical
information systems enable feature-based queries, like finding port cities or nearby
restaurants using mobile apps.
Design or Diagram Search :Design tools provide limited search capabilities within
collections. Advanced strategies can identify specific measurements, and document-
structural recognition aids in searching for specific layouts.
 Sound Search :Music-information retrieval (MIR) systems enable querying via audio input,
such as singing a melody. Recognizing individual performers is improving, but finding
spoken phrases in audio remains challenging.
Video Search:Video searches require segmenting content into scenes for easier
navigation. Projects like Info-media use visual and textual features to enhance video
retrievability.
Animation Search:Animation authoring tools are advancing, allowing searches for specific
animation types, such as spinning globes or morphing faces.

Advanced Filtering and Searching Interfaces

Here are the key alternatives to form-fill-in query interfaces summarized in short points:

1**Complex Boolean Queries**:

- Allow intricate searches but can confuse users.

- Misinterpretation of "and" and "or" is common.

2. **Automatic Filtering**:

- Users create keyword profiles for real-time document filtering.

- Alerts users to relevant new information.

3. **Dynamic Queries**:

- Interactive sliders/buttons for adjusting queries in real-time.

- Reduces syntax errors and encourages exploration.

4. **Faceted Metadata Search**:

- Combines category browsing with keyword searching.

- Allows users to refine searches across multiple dimensions.

5. **Collaborative Filtering**:

- Uses user ratings to recommend items based on collective preferences.

- Helps users discover new content.

6. **Multilingual Searches**:

- Users can select dictionaries and customize translations.

- Enhances accuracy beyond basic translation tools.

7. **Visual Searches**:
 - Utilizes graphical representations for easier query specification.

- Improves context and reduces user errors.

Information Visualization:
Introduction
Direct-manipulation interfaces use visuals like maps and photos for better user
understanding. Advances in technology enhance how large datasets are displayed and
manipulated.

Information visualization helps users identify patterns in abstract data through interactive
graphics. Unlike scientific visualization, it focuses on categorical data.

Some users prefer text-based interfaces for their familiarity and efficiency. Visual tools must
clearly benefit users and be accessible to all.

The "visual-information-seeking mantra" guides effective data navigation. It suggests starting

with an overview, then zooming in and filtering before showing details.

Data types of task taxonomy

- Emphasizes the principle of "overview first, zoom and filter, then details on demand" for
effective exploration.

- Uses a data type by task taxonomy to identify user needs and visualization challenges
across various data structures.

1.One-Dimensional Data

-Comprises linear formats like program source code and text documents organized
sequentially.

- Allows interaction with attributes such as modification dates through color-coding and
design features.

2.Two-Dimensional Data

- Includes geographic maps and floor plans, where each item has specific spatial attributes.

- Enables tasks like finding adjacent items and completing basic visualization activities.

3.Three-Dimensional Data
- Represents real-world objects and complex relationships, applicable in fields like medical
imaging.

- Requires users to navigate volumes and surfaces while managing potential occlusion issues.

4.Multidimensional Data

- Allows manipulation of items as points in multidimensional space, often using dynamic

scatter plots.

- Faces challenges like occlusion and disorientation in three-dimensional visualizations,

highlighting the need for effective design.

5.Temporal Data

- Temporal data represents time series, where events have distinct start and finish times,
allowing for overlapping occurrences.

- Common tasks include finding events within specific time periods and comparing periodic
phenomena, supported by various project management tools and visualizations like the
Perspective Wall.

6.Tree Data

- Tree data consists of hierarchies where each item (except the root) links to one parent,
allowing for multiple attributes on both items and links.

- Users can apply basic tasks to items and explore structural properties, such as depth and
the number of subordinates in an organizational chart.

7.Network Data

- Network data involves complex relationships among items, where users seek to find the
shortest or least costly paths between them.

- Visualization methods include node-and-link diagrams and matrices, although effective

network visualization remains challenging due to the intricate nature of user tasks and
relationships.

The seven essential tasks for effective information visualization are:

1. **Overview Task**: Provides a broad view of the entire dataset to understand overall
structure.

2. **Zoom Task**: Allows users to zoom in on specific items while keeping context.

3. **Filter Task**: Users can remove uninteresting items to focus on relevant data.
4. **Details-on-Demand Task**: Users can access detailed information about selected items
easily.

5. **Relate Task**: Visuals help users see relationships between data points, like proximity or
color coding.

6. **History Task**: Tracks user actions for features like undo and replay, aiding exploration.

7. **Extract Task**: Enables users to save or share selected data subsets for further use.

These tasks help design user-friendly interfaces for navigating and analyzing complex data.

Challenges for information visualization

The data type by task taxonomy outlines several challenges in creating effective information
visualization tools:

1. **Import Data**: Organizing and formatting data can be complex, requiring filtering and
normalizing values.

2. **Combine Visuals with Text**: Visuals need clear labels that enhance understanding
without cluttering the display.

3. **See Related Information**: Users often need access to related data to make informed
decisions, requiring quick access to context.

4. **View Large Volumes of Data**: Many tools struggle to handle large datasets interactively,
making it hard to visualize millions of items.

. **Integrate Data Mining**: Combining visualization with data mining helps identify patterns
and trends.

6. **Collaborate with Others**: Effective discovery often relies on teamwork, so tools should
allow easy sharing of findings and insights.

7. **Achieve Universal Usability**: Making tools accessible for all users, including those with
disabilities, is essential and challenging.

Addressing these challenges can improve the effectiveness of information visualization tools
for everyone.