GoogleSearch

Description

GoogleSearch is a library of simple R functions and scripts (Mazanec, 2008) to launch automated Google queries for capturing occurrence and co-occurrence frequencies of keywords. Two examples taken from a city tourism application are provided. In these examples city names denote the rows and attributes define the columns of the co-occurrence matrix. A routine for non-disjunctive hierarchical clustering (Peay, 1975) is also provided. See the Run... scripts for detailed working steps.

A second demo script tailored for the www.tripadvisor.com site is also available (JM, 10/2008). It captures the frequencies reported under the 'What Our Users Are Saying Results ...'  heading. For small destinations and very specific attributes the limited search domain of TripAdvisor may deliver zero occurrences or data too sparse for hierarchical clustering and/or multidimensional scaling.

Note that keywords must be in English and composite words are set between quotation marks ("New York", "peace and quiet").

Literature

One of the functions is based on Cilibrasi, R. L. and M. B. Vitányi (2007) The Google Similarity Distance, IEEE Transactions on Knowledge and Data Engineering, 19 (3), 370-383.

For CLIP clustering see Peay, E. R. (1975) Nonmetric Grouping: Clusters and Cliques, Psychometrika, 40, 297-313, Mazanec, J. A. (1978) Strukturmodelle des Konsumverhaltens, Vienna: Orac, pp. 447-453, and Mazanec. J. A. (1997) International City Tourism, Analysis and Strategy, London & Washington: Pinter, pp. 240-242.

The list of positive connotations in example #2 below is based on the System for Connotative Analysis of Discourse (US Patent of December 18, 2001) by Wayne O. Chase.

Chapter 10 of Mazanec, J. A. and K. W. Woeber (2009), Analysing International City Tourism, Vienna: Springer, pp. 191-210 presents a fully elaborated case study for tourist cities.

A study on receiving countries is forthcoming in the Journal of Travel Research (Mazanec, J. A., Tourism Receiving Countries in Connotative Google Space).

Requirements

You need the open source computing environment R which you may freely download from http://cran.r-project.org/.

Download

Please read and accept our download conditions before!

Download and unzip one of the zip files to a separate directory of your choice and make it your R working directory. Then start R, edit the RunGoogleS or RunTripAdvS script and execute steps 1-5 sequentially. Submitting 58 connotative search items for 23 cities may take a while.

File

Version

OS

Location

Size

demo_6_4.zip contains the R scripts, CLIP50.exe and demo data for 6 cities and 4 arbitrary search items

 1.2 (08/2009)

 Windows XP, Vista;

 R 2.7.0 or higher

HTTP

 64 KB

demo_23_58.zip contains the R scripts, CLIP50.exe and data for 23 cities and 58 connotative search items

 1.2 (08/2009)

 Windows XP, Vista; 

 R 2.7.0 or higher

HTTP

 71 KB

tripadv.zip contains the R scripts, Clip50.exe and demo data for 6 cities and 4 arbitrary search items for searching at www.tripadvisor.com

 1.2 (08/2009)

 Windows XP, Vista;

 R 2.7.0 or higher

HTTP

 64 KB