Anahtar kelime tabanlı arama motoru optimizasyonu kelime çantası algoritması kullanarak

Text data for analysis is an essential activity in science. The diversity of scientific fields produces very heterogeneous data in terms of type and information provided. Many projects have been carried out in our laboratory to extract semantic elements from images from different backgrounds Document classification or document categorization is a problem in library science, information science and computer science. This task consists of assigning one or more classes or categories to a document. This can be done "manually" (or "intellectually") or through algorithms The main contributions of this thesis is Creating a new way of interacting with the user, much more personalized. the data we collect today allows us to see things that until recently were too big for us to see". data are categorized, that is, if keywords, which indicate the main content of the data, were associated with these data. Scientific articles, for example, when associated with keywords, help in the analysis of the profile of the conference in that year or of a journal issue.in this work we used the bag of words algorithm for keyword extraction from the image and search related content in the database. The BOW features are way better than RGB and HSV when it comes to color based extraction as it is developed as a further iteration to these two. Also the EHD proves to be a better method than Wavelet and other segmentation methods. There are certain anomalies however that arise due to the combination of these two features. Too much of a color in the background takes away the weightage of the centre object. Such problems exist in every CBIR system but with further analysis we can surely work on improvement.