Bing Pan

1
Travel Information Search on the Internet An
Exploratory Study
Bing Pan National Laboratory for Tourism and
eCommerce Department of Leisure
Studies University of Illinois at
Urbana-Champaign Advisor Dr. Daniel R.
Fesenmaier
2
Introduction

• Overall structure
• Problems/background
• Literature review
• Conceptual framework
• Methodology
• Results and conclusions
• Implications

3
Problems/Background

• Tremendous development of the Internet and widely
  use of the Internet as a source for travel
  information
• Information overload
• Finding relevant information
• The vocabulary problem (Furnas, Landauer, Gomez
  Dumais, 1987 )
• Understanding of the users is essential!
• Research on travel information search is scarce
  most are survey research
• Adolescence era of information technology
• The goal of Man-Computer Symbiosys (Licklider,
  1960)

4
Travel Planning and Information Search

• Travel planning is (Jeng, 1999)
• A hierarchy of decisions which involves a set of
  sub-decisions, for example, destination, travel
  partners, accommodation, dining and others
• Multi-facet, dynamic and contingent process.
• People use different combinations of information
  sources searching for travel information (Fodness
  Murray, 1998)
• The use of different sources depends on the cost
  and benefit of accessing certain information
  sources (Ratchford, Talukdar, Lee, 2001)

5
Information Seeking and the Internet

• Information search on the Internet as navigation
  through hypertext
• Navigational task vs. information task (Kim and
  Hirtle, 1995 )
• Information search experience and domain
  knowledge contribute to successful and
  satisfactory information search (Hsieh-Yee, 2001)

6
Information Seeking and the Internet

• Information Foraging (Pirolli Card, 1999)
• Information searchers use proximal cues to
  identify important information for further
  exploration or consumption
• Information Scent (Chi, Pirolli, Chen and
  Pitkow, 2001)
• Information searchers identify valuable
  information from the snippets of proximal cues

7
Mental Models and Information Search

• Mental model as intermediate construct
• Semantic mental model vs. mental model in HCI
• Declarative Knowledge vs. Procedural Knowledge
  (Anderson, 2000)
• Using semantic networks to represent semantic
  mental models (Collins and Quillian, 1972
  Doerfel, 1998)

8
Mental Models and Information Search

• The mismatch between users mental model and
  systems conceptual model contribute to usability
  problem (Norman, 1986)
• The polyrepresentation of concepts in the
  languages between the users cognitive space and
  the information system is a major issue when
  designing an effective information system
  interface (Ingwersen, 1996).
• Mismatch between travel information searchers
  mental model and the semantic model of travel
  information space..

9
A Conceptual Model of Travel Information Search
on the Internet

• Travelers semantic mental model includes
  destination, travel partners, accommodation,
  dining and others
• Choices of links are based on the relative value
  of information scent the link anchors (texts or
  pictures)
• Navigation process and reading process
• Search process can be broken up into different
  episodes each episode targets at a sub-problem
• Travel experience, computer and Internet
  experience, and the experience of using the
  Internet as travel information source, contribute
  to the congruence of mental model between
  travelers and the conceptual model of information
  space and the further congruence of models will
  contribute to the satisfactory travel information
  search/travel planning process.

10
Semantic Mental Model
Destination
Tropical
Disneyland
South
Theme Parks
Activities
Museums
Shopping
Exciting
Art
Scuba Diving
Boating
Florida
11
Conceptual Model
Common Concepts
12
Main Goals of Research

• Understand the process and the structure of
  travel information search on the Internet
• Discover the congruence and discrepancies of
  semantic mental model of travel information
  searchers and the semantic model of tourism
  information providers
• Examine the extent to which the discrepancy of
  semantic models of travelers and tourism
  information providers contributes to the
  satisfaction of information search on the
  Internet and travel experience, computer and
  Internet experience, and the experience of using
  the Internet as travel information source,
  contribute to the congruence of the two semantic
  models

13
Research Procedure

• Pilot study (May - June, 2002)
• 5 subjects
• Formal experiment (September, 2002)
• - 15 diverse subjects (Eveland, Dunwoody, 2000
  )
• Sources for measuring consistency
• Websites visited
• Traveler's semantic models
• Travel experience and Internet use experience

14
Research Methodology

• An travel planning experiment on the Internet
  regarding a weekend trip to a designated
  destination (San Diego, CA)
• Setting
• 1. Using a web browser on a workstation located
  in the National Laboratory for Tourism and
  eCommerce
• 2. A tester will carry out the pre-experiment
  survey, interviews, experiments, and
  post-experiment questionnaire.
• Tasks
• 1. Choose/identify their activities and
  accommodation during their 2 day vacation to San
  Diego, CA.
• 2. Need to write a short essay regarding their
  travel plan
• 3. One-hour is the maximum length of travel
  planning
• 4. One subject will be randomly selected
  to win two round-trip airline tickets

15

Research Procedure
Travel experience, computer and Internet use
experience, the experience of using Internet as
travel information source
Research Procedure
Goals
Phase I Analysis of the process
I. Pre-experiment Survey
Obtain 1. Demographical variables 2. Travel
experience 3. Computer and Internet use
experience 4. The experience of using the
Internet for travel planning purpose
SNA on transcripts of interviews
Phase II Comparison of two semantic models
II. Pre-experiment Interview
Obtain travelers initial semantic mental model
regarding one destination prior to travel planning
III. Travel Planning Experiment
Obtain information search protocol in order to
1. Understand the process of travel planning on
the Internet 2. Explore tourism information
space regarding one destination
Protocol Data
SNA on web pages
Phase III Modeling Satisfaction
Satisfaction
IV. Satisfaction Survey
Obtain satisfaction (including process
satisfaction and outcome satisfaction)
Reasons for satisfaction
V. Post-experiment Interview
Understand the Internet as travel information
source understand the satisfaction of travel
information source clarify the ambiguities of
information processing.
16
Research Methods

• Protocol analysis using transcriptions of
  information search behavior including information
  search activities, information processing, and
  computer use activities
• Semantic network analysis using transcripts from
  interviews and text from full texts of web pages
• Transcripts from interviews
• Full texts of web pages from visited web sites
• Correlation analysis using survey data and
  results from semantic analysis

17
Phase I Results

• 10 undergraduate students, 1 graduate student,
  and 4 researchers.
• 19 to 45 years old with an average age of 25
  years.
• 6 subjects were men and 9 women.
• 4 have been to San Diego at least once.
• Most of them have used the Internet to check out
  destination, hotel and airline information.
• Used computer from 4 to 28 years with an average
  of 14.3 years. They used the World Wide Web and
  email from 4 to 17 years with an average of 7.6
  years.
• An average of planning time of 36 minutes, with
  minimum time of 20 minutes and a maximum time of
  55 minutes.
• Visited from 7 to 26 web sites with an average of
  15. Visited from 60 to 312 web pages, averaging
  124 web pages.
• 7 of 15 used a printer to print out information
  and organize information 12 used a piece of
  paper 3 used wordPad or Microsoft Word.

18
Phase I Information Search Protocol Data
19
Research Procedure
20
Phase I Information Search Protocol Data
21
Research Procedure
22
Phase I Information Search Protocol Data
23
Phase I Final Protocol
24
Understanding of the Process
A Click Semantic Map
25
Phase I Further Results

• Internet is always used with other software and
  information organization and decision aid tools
  (other software, paper and printer).
• Travel information search can be broken up into
  different episodes, in which each episode may
  encompass different web sites.
• The model of navigation and reading process is
  too simple. The informational behavior involves
  searching, browsing, navigating, reading and
  information organization.
• Information hubs are frequently used in the
  process. A information hub is a web page leading
  to a cluster of relevant web pages. People rely
  on information hubs to minimize inter-cluster
  information seeking cost and they are reluctant
  to change information hubs.
• Experienced Internet users tend to use multiple
  clusters when searching for information.
• Geographical information and time frame are not
  well represented. Travelers need information
  organization and travel planning in their minds
  or using decision aids they need to switch
  between different attractions and online map web
  sites.
• Certain places and attractions are not
  searchable, for example, hotel facing the sea,
  bars on the beach and etc.

26
Phase II. Tourism Information Space
27
Phase II. Comparison of Two Semantic Models
Tourism Information Space
Travelers Mental Model
28
Phase II. Semantic Model of Tourism Information
Space
29
Phase II. Travelers Semantic Mental Model
30
Phase II. Comparison of Two Semantic Models
Located in California, San Diego City is famous
for its tours of harbor and Sea World. Here you
can find discount information on cruises,
tickets, various services, and maps. You can go
to a shopping center, a park, or a free event. We
have many best attractions toward different age
groups with reasonable price, such as missions,
hospital buildings, and bay area. Most of them
are close to bus stations. We have many famous
museums, most of them are located at Balboa Park,
including art museum.
We know San Diego is a famous tourist city
located in California. If I go there, Id like to
walk around the beaches and the city in the day.
Id like to watch people, see the little things
around the city. Id like to stay at good and
nice hotels and places, dining at good and nice
restaurants. I also like to try different types
of food. Id like to look for attraction
information, try to look for interesting museums.
I also like to visit the big thing in San Diego,
like the zoo. During the night Id like to go to
music clubs.
31
Conclusions

• Internet is used by travel information searchers
  in different ways and combined with different
  decision aids
• Travel information searchers use large quantity
  of web sites to search for tourism information
  and the information space they encountered are
  very diversified.
• The process of travel information search on the
  Internet can be broken up into different episodes
  and each episode solves a sub-problem.
• Information hubs are frequently used by
  searchers to minimize the inter-cluster
  information search cost.
• Geographical information and time-constraints are
  two important information which obtruded the
  efficient travel planning on the Internet.
• Semantics posed an important discrepancy between
  travelers and tourism information online. From a
  marketers perspective, online tourism
  information is too marketing-oriented and ignored
  the language of travelers.

32
Implications

• Considering different use of the Internet
• Use travelers language
• More collaborations are needed between different
  groups of a destination
• Provide decision aids with geographical
  information and time frame
• Provide meaningful tourism ontologies for
  semantic webs and semantic data modeling
• Dominance of language/shift of power
• Provide novel and exciting information vs.
  fulfill users current goals

33
Comments and Questions?