Volume XLII-5
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-5, 189–192, 2018
https://doi.org/10.5194/isprs-archives-XLII-5-189-2018
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-5, 189–192, 2018
https://doi.org/10.5194/isprs-archives-XLII-5-189-2018
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

  19 Nov 2018

19 Nov 2018

LOCATION BASED ADVERTISING FOR MASS MARKETING

S. Vignesh Kandasamy, A. Madhu, P. K. Gupta, A. Niveditha, and K. Bordoloi S. Vignesh Kandasamy et al.
  • Geoinformatics Department, Indian Institute of Remote Sensing, Indian Space Research Organisation, Department of Space, Dehradun, Uttarakhand, India

Keywords: Location based Advertising, Open Source GIS, Tesseract OCR

Abstract. GIS and machine learning (ML) are powerful ICT tools in retail industry which helps the sellers understand their markets. For the consumers, however, there always lies an ambiguity with respect to the quality and quantity of the product to be purchased, vis-à-vis the price paid for it. Most retail businesses today adopt “Discount Pricing Strategies” or “Offers” to make new customers and increase sales. Owing to several establishments selling the same product and offering a variety of offers, the process of identifying the shops where the consumer can get the best value for his money, requires a lot of manual effort. A prototype has been developed in this study to allow the consumers to locate such prospective shops based on advertisements in newspapers. This solution has a two-pronged approach. First, all the offers advertised in the newspaper are pre-processed and text extraction is performed using a ML algorithm named Tesseract OCR. Second the location of shops is collected and stored in a geodatabase. Finally, the advertisement is matched to the respective geo-located shop based on its name and location. Further based on the location of the consumer and his purchase choice, shops offering discounts are shown on a web based map. This prototype provides the consumer, a platform for geo-discovery of establishments of interest through the clutter of unrelated endorsements, by the use of Open Source GIS, Python programming and ML techniques.