From query analysis to user information needs: a study of campus map searches
| Pages | 104-129 |
| DOI | https://doi.org/10.1108/LHT-12-2014-0110 |
| Date | 21 March 2016 |
| Published date | 21 March 2016 |
| Author | Zoe Chao |
| Subject Matter | Library & information science,Librarianship/library management,Library technology |
From query analysis to
user information needs: a study
of campus map searches
Zoe Chao
Penn State University, University Park, Pennsylvania, USA
Abstract
Purpose –Search engines and web applications have evolved to be more tailored toward individual
user’s needs, including the individual’s personal preferences and geographic location. By integrating
the free Google Maps Application Program Interface with locally stored metadata, the author created
an interactive map search for users to locate, and navigate to, destinations on the University of
New Mexico (UNM) campus. The purpose of this paper is to identify the characteristics of UNM map
search queries, the options and prioritization of the metadata augmentation, and the usefulness and
possible improvement of the interface.
Design/methodology/approach –Queries, search date/time, and the number of results found were
logged and examined. Queries’search frequency and characteristics were analyzed and categorized.
Findings –From November 1, 2012 to September 15, 2013, the author had a total 14,097 visits to the
SearchUNM Maps page(http://search.unm.edu/maps/).There were total 5,868 searches (41 percentof all
the page visits),and out of all the search instances, 2,297 of them (39 percent)did not retrieve any results.
By analyzingthe failed queries, the authorwas able to develop a strategyto increase successful searches.
Originality/value –Many academic institutions have implemented interactive map searches for
users to find locations and navigate on campus. However, to date there is no related research on how
users conduct their searches in such a scope. Based on the query analysis, this paper identifies user’s
search behavior and discusses the strategies of improving searches results of campus interactive maps.
Keywords Information retrieval, User interfaces, Data analysis, Computer applications,
Geographic information systems, Information searches
Paper type Research paper
Introduction
Students often come to a library’s reference desk to ask for directions to a specific
building or classroom (Bishop, 2011). Typically, they will be shown a print map and
offered a verbal explanation, helping them to their destination. Yet, wayfinding
interfaces on the web have evolved dramatically; search engines like Google and
Microsoft Bing have developed interactive maps for users to locate themselves in
real-time and made them freely available as smart phone apps. The prevalent use of
smart phones among college students underscores a need for institutions to create truly
digital campus maps in order to provide non-static, interactive location searches.
To understand a user’s geographic information (henceforth denoted geo-info) need, it is
important first to define what geo-info is. Raper (2007) described geo-info “as information
whose representation and communication is dependent on the constitutive role of place,
space, and time.”In principle, geo-info can be represented and communicated in either
geo-centric coordinate referencing or places and landmarks (Raper, 2007). For example,
Google Maps uses maps, places, imagery, directions, and distance/time to represent the
information and to communicate results with users. In the past decade, researchers have
Library Hi Tech
Vol. 34 No. 1, 2016
pp. 104-129
©Emerald Group Publishing Limited
0737-8831
DOI 10.1108/LHT-12-2014-0110
Received 1 December 2014
Revised 23 October 2015
Accepted 29 October 2015
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0737-8831.htm
Special thanks go to Randy Johnson and Christin Sraha at the University of New Mexico for their
hard work on this project.
104
LHT
34,1
proposed the concept of geographic relevance (GR) to address the spatial-temporal context
in geographic information retrieval. Raper (2007) defines GR as “a spatial-temporally
extended relation between information need and geographic information object.”In order to
be geographic relevant, i.e. to meet user’s geo-info need, three aspects need to be considered.
First, a geographic entity in the real world must be well represented in a computer system.
Second, a user’s spatial-temporal context, e.g. location, mobility, etc., must be taken into
account, in addition to the normal relevance ranking. Third, the interface must be able to
communicate the relevant geographic information effectively (Raper, 2007; De Sabbata and
Reichenbacher, 2012).
In 2012, the campus Information Technologies (IT) department and the University
Libraries (UL) at the University of New Mexico (UNM) created an interactive map
search for users to locate, and navigate to, destinations on the UNM campus by
integrating the free Google Maps Application Program Interface (API) with locally
stored metadata. The Google Maps API has been applied to provide geo-info in many
different contexts: to optimize resource distribution (Wang et al., 2009); to refer
hospitals (Kobayashi et al., 2010); and to provide real-time traffic flow (Wu et al.,
2007). It is also more light-weight and less burdensome, compared to OpenLayers
API and ArcGIS API (Fernandes et al., 2013). More importantly, it is an economically
sound choice (Pan et al., 2007; Kobayashi et al., 2010). We made the assumption that
users reveal their current geo-info interest and information need by submitting
a query to a search engine. By identifying what people search for and the words used
in such searches, we hoped to gain a better understanding of the interaction between
users and our map search system, as well as to evaluate the representation
of geographic entities on campus, the impact of user’s context, and the effectiveness
of our interface.
The research questions we hope to answer are the following:
RQ1. What are the characteristics of a UNM map search query?
RQ2. Do the data and metadata of the geographic entities on campus align with the
user’s information needs?
RQ3. Does our map search interface communicate effectively enough for users to
retrieve the information?
Literature review
To retrieve information from general web resources, the classic information retrieval
system (CIRS) generally evaluates relevance based on the relation between queries and
the text in web documents. If often does not consider a user’s spatial-temporal context,
i.e. GR. To facilitate the geo-info retrieval in a CIRS, researchers have studied the
characteristics of geographic queries (henceforth denoted geo-query), the geographical
content of web resources, and the indexing and processing power of the search tools.
Additionally, there is research focussing on systems primarily designed for geo-info
known as geographic information retrieval systems (GIRS), for example, a Google-
Maps-based traffic information system.
Though two different systems, CIRS and GIRS are not exclusive of each other, as a
matter of practical use. Often, CIRS can serve as a gateway to GIRS: for example, a user
does a Google search (interacts with a CIRS) and clicks a link to Google Maps
(interacts with a GIRS). In many cases, however, the connections between GIRS
and CIRS are absent. As a result, the rich geo-info developed in GIRS is not exposed
105
A study
of campus
map searches
To continue reading
Request your trial