Scrappy Job Search  (aka The Job-Difference Search Engine)


Basic Options

Links: Vet Search  Home

Advanced Options

Derives differentiating terms from complete documents not search snippets. This involves download overhead plus extra term analysis.
Start with textual view;   Use no images in traditional view.
Target dimension:          

What difference does it make?

Tired of the same old textual search snippets? The Job-Difference Search Engine (aka Visual "Diff" Search) provides a new way of presenting search results via images and differentiating terms, with titles included as well. This format makes it much easier for you to scan the search results to find items of interest compared to traditional search snippets. For more details of a particular item, just hover the mouse over the info icon for that cell (visible when mouse is over the cell). For an illustration, see the popup snippet in the bottom right corner of the example results in the following Help section.


Enter the search query using keywords just as you would for Google, and then click search or press the enter key. (You can also use special operators, such as 'keyword' and 'site:name'.) The result will be in a tabular format, as in the example below, where each image serves as a convenient handle for moving, resizing, etc. Click the info icon in upper left corner to see the traditional search snippet. The popup snippet will change when a different cell is hovered over, and it will go away after a few seconds when the mouse is moved away from the cell. Click the pin icon to keep the snippet always visible.

Each cell can be resized, moved or deleted independent of the others. To resize, click-and-drag from the lower righthand corner. To move, click-and-drag from the image. To delete, click the x icon in the upper righthand corner. The following example is the result of such cell modifications, along with the hovering mentioned above.

night owl visual search example (after editing)

The fine print

note: This is a beta release that only supports a limited number of searches, due to the cost of search calls and to the computing infrastructure.

Copyright (c) 2017-2022 ScrappyCito, LLC
Patent Allowance: US16/041,730

Other copyrights:
     Portions copyright (c) 1998-2022 Google, Inc.
     Portions copyright (c) 2004-2022 Indeed, Inc.
     Portions copyright (c) 2003-2022 SciPy
     Portions copyright (c) 2003-2022 NumPy
     Portions copyright (c) 2007-2022 scikit-learn
     Portions copyright (c) 2016 elzilrac (e.g., tf-idf package)
     Portions copyright (c) 2015 shyyko.serhiy (e.g., Rulez.js for debugging)
     Portions copyright (c) 2022 (categorization images)
     Portions copyright (c) 2022 (categorization images)
     Portions copyright (c) 2022 (categorization images)
     Portions copyright (c) 2022 (gallery and related images)
     Portions copyright (c) 2009-2022 Radim Řehůřek (e.g., gensim)
     Portions copyright (c) 2016 yrq110 (e.g., sample Google API usage)
     Portions copyright (c) 2001-2022 Wikimedia Foundation (e.g., Wikipedia for text categorization)
     Portions copyleft (ɔ) 1983-2022 GNU (e.g., invaluable GNU/Linux toolset used during development)

Modified source and data (as per licensing of above):


¡Muchas gracias to our supporters!

Send comments about this page.