Java Google Query

the project is a java software it read text files CSV and output CSV it use one library for text categorisation/classification feature, the library read a sentence and return its probable category, it need a training process before to be able to recognize categories correctly Project is composed of the following module Module google-query - query google with a text ( the list of texts query are provided in text file) - parse google first 100 results - load all 100 results pages Module Analyse-Page ( analyse each page loaded by module google-query ) - with jsoup split the page into paragraph text - analyse each paragraph with the Paragraph-Analyser - analyse all the paragraph results with the Page-Paragraphs-Analyser Module Text entity recognition - use [login to view URL] or [login to view URL] or better - we provide list of street name, people name, cities names, and some items-categories/groups lists to train model - In the text paragraph , All entity recognized should be replaced by their entity name Module Paragraph-Analyser trainer - use [login to view URL] or [login to view URL] or better - We provide a list of 1000 paragraph texts with the category/classification - the trainer will build a model from the 1000 sample to use by Module Paragraph-Analyser Module Paragraph-Analyser - take a paragraph text as input and give a probable detected category/classification - use library [login to view URL] or [login to view URL] or better Module Page-Paragraphs-Analyser trainer - We provide 1000 pages URL and the category/classification of each pages-url - we need to load the url and to process it with Module Paragraph-Analyser - the Page-Paragraphs-Analyser receive the Paragraph-Analyser result and is trained to match the provided category/classification - use [login to view URL] or [login to view URL] or better Module Page-Paragraphs-Analyser - receive category/classification of all paragraphs of the page, then it concatenate all into a single text string - give a probable global category/classification - as input is a text string composed of multiple words we use the same library : [login to view URL] or [login to view URL] or better note: We call a page : a HTML page loaded from a URL We call a paragraph : a text inside a HTML-Block-Tag like <DIV>, <TD>, <H1> all others tags like <A>,<P>,<B>,<SPAN> are stripped out from the block , only text remain We call a category/classification : a simple text-name (color, city, etc..) We call a Text entity : a text (example green, dark red, yellow) which have a entity name (example colors)