A Modified Approach for Automatic Extraction of Geographic Data
Ammar Sikander, Ying Zhang, Usman Anwar, Adeel Sikander · 2019 2nd International Conference on Computing, Mathematics and Engineering Technologies (iCoMET) · 2019
The popularity and growing of Big Data are increasing day by day, but the data is congested and unstructured. Finding of sorted and precise data according to requirements is not an easy task. In the era of social networking & machine learning, researchers have significant interests to retrieve automatic extraction of desired information. The focus of this scrutiny is to combine different techniques with their best results based on divisible tasks like regular expression, pattern based, machine learning as a classification problem and logic-based approach. This study is eventually leading to precisely identifying the locations. We extract the data of multiple blocks of the United States from the open web. After that, classification technique is applied with a machine learning approach by using SVM to get the specific information according to requirements. After getting the first phase results, the particular names are searched with their places from search engines to confirm the locations. The outcomes, finally compared with an Open Street Map (OSM) and Wikimapia. The results have shown the excellent approximation to google maps and more precise than OSM and Wikimapia.