“I Prefer Regular Visitors to Answer My Questions”: Users’ Desired Experiential Background of Contributors for Location-based Crowdsourcing Platform

Title of the Paper

“I Prefer Regular Visitors to Answer My Questions”: Users’ Desired Experiential Background of Contributors for Location-based Crowdsourcing Platform

Paper Information

• Domain: Human-Computer Interaction and Location-based Mobile Crowdsourcing Platforms
• Keywords: Mobile Crowdsourcing, Location-based, Information Needs, Information Quality, Contributor Experience, Task Allocation

Research Background and Issues

• Identified Problems or Challenges:

  • Information on location-based crowdsourcing platforms comes from diverse contributors, but current platforms lack contextual information about contributors' experiential backgrounds, which can confuse users when faced with diverse reviews.
  • There is no clear research on which types of contributor experiences can enhance the utility of descriptive information.

• Significance:

  • Users may prefer information provided by contributors with specific backgrounds, which is critical for platforms to customize task allocation and improve user satisfaction.

• Research Motivation and Related Work:

  • From the perspective of matching contributors with tasks, existing research focuses on matching based on skills, location, or cognitive abilities but has not sufficiently explored the relationship between contributors' experiential backgrounds and the utility of information in location-based contexts.

Research Questions

This study addresses the following three core questions:

1. What types of information do users seek on location-based crowdsourcing platforms, and what are the characteristics of this information?
2. Which specific aspects of descriptive quality do users value the most?
3. What contributor experiential backgrounds do users believe enhance the utility of descriptions?

Proposed Solution

• Research Methods:

  • A three-phase research approach was employed, including:
    1. Semi-structured Interviews: To explore users’ initial perspectives on the quality of location-based information and contributors’ experiential backgrounds.
    2. Survey: To identify the 8 most commonly searched types of location-based information and their characteristics.
    3. Online Scenario Experiment: To quantify the relationships between information types, descriptive quality attributes, and contributors’ experiential backgrounds.

• Innovations:

  • Proposed a new framework for matching contributors’ backgrounds with users’ needs for descriptive quality.
  • Clarified the unique value of contributors’ experiential backgrounds for different types of information.

Research Findings

• Specific Findings:

  1. Interviews revealed five core characteristics of location-based information (e.g., objectivity, relativity) and ten key aspects of descriptive quality (e.g., completeness, timeliness).
  2. The survey identified 8 most commonly searched types of information, categorized into three groups based on their attributes.
  3. The online experiment highlighted universally beneficial factors in contributors’ experiential backgrounds (e.g., recency, visit frequency) and specific background preferences for certain types of information.

• Advantages Over Existing Solutions:

  • Introduced a dimension of contributors’ experiential backgrounds based on user preferences into existing task allocation mechanisms.
  • Provided user-centric task allocation strategies to improve information relevance.

• Experimental or Evaluation Results:

  • Universally Applicable Contributor Backgrounds: Recency and visit frequency were deemed universally applicable across all information types.
  • Type-specific Preferences: For example, users preferred long-term residents to answer parking-related questions but favored diverse visitation experiences for scenic and recommendation-related information.
  • Relationship Between Descriptive Quality Attributes and Contributor Backgrounds: For instance, time specificity predicted the importance of long-term residency experience.

• Limitations and Future Directions:

  1. The scope of information types may not fully encompass all real-world platform needs; future research should expand to different cultural and geographical contexts.
  2. Gender distribution bias (overrepresentation of female participants) requires further testing for gender applicability of the results.
  3. The study did not directly investigate the relationships between information attributes, descriptive quality, and backgrounds; future research could supplement this analysis.

In summary, this study provides valuable insights for designing task allocation mechanisms based on contributors’ experiential backgrounds, while emphasizing the importance of shifting from a “local” to an “experiential” perspective in location-based crowdsourcing platforms.