Through “Web-harvesting,” Jones et al.’s Modelling Vague Places (2008) introduces techniques to improve modeling vague places (1048). I was interested in how Jones et al. utilized “place names” from the Web to create their models because I am following a similar methodology for my own research. While researching for my own project on volunteered geographic information (VGI) and Twitter harvesting, I read an article by Elwood et al. (2013) called Prospects for VGI Research and the Emerging Fourth Paradigm that explains how people have a tendency to use place over space when they contribute geographic information through a public platform (i.e. social media or blogs). For example: a Tweeter may post a street-name without geotagging their post, thus the only geographic information they are providing is a place attribute, not any coordinate information. This makes it more difficult to gather specific/precise spatial information when crowd-sourcing data from the Web. What is similar to my project’s methodology and Jones et al.’s article is that we both look at “semantic components” (Bordogna et al. 2014, p. 315), meaning we both are identifying textual Web information to gather information on the “precise places that lie within the extent of the vague places” (1046). Additionally, Jones et al. “decrease[d] the level of noise” through filters, something I also will be doing while harvesting Tweets (1051). With comparable methodological approaches, I will certainly consider some of Jones et al.’s techniques while completing my own project.
Similarly to what we discussed last class, this article also highlights issues with ‘big data;’ specifically, how can we sift through so much heterogeneous data and pull out the most relevant information in an efficient and time-saving approach? Jones et al. introduce strategies to sift through the Web’s big data, but it would be interesting to see how these techniques have changed within the past 7 years since this article was published. CyberGIS could certainly improve the validity of gathering “published texts” off the Web through solving technological issues, such as improving automated algorithms that affected the results of Jones et al.’s research (1048).
One final point, the digital divide was not mentioned within this article. Although Jones et al. focused their research only within the U.K. where a richer demography have the capabilities to access the Web, it is important to consider that local people from a poorer locality may not be providing any information to the Web. This ignores local people’s interpretations of their landscape/place, which would be considered “rich in geographical content” if they could contribute information to he Web (1051).
-MTM
Bordogna, G., Carrara, P., Criscuolo, L., Pepe, M., and Rampini, A. (2014). A Linguistic Decision Making Approach to Assess the Quality of Volunteer Geographic Information for Citizen Science. In Information Science, 258, 312-327.
Elwood, S., Goodchild, M., Sui, D. (2013). Prospects for VGI Research and the Emerging Fourth Paradigm. Editors D. Sui, S. Elwood, M. Goodchild, Crowdsourcing Geographic Knowledge: Volunteered Geographic Information (VGI) in Theory and Practice (361-376). Dordrecht: Springer.