Streamline visual identity of "search" function in contract view to annotation

[anderspeders]

Could this in visual terms appear a bit more like the annotations - eg. by using the same round dots for each search result.

[KaitlinCCSI]

@anderspeders - not sure what you mean by round dots, but #373 asks for green dots to be removed if they do not serve a function. @jedm, any thoughts?

[byndcivilization]

@KaitlinCCSI I think we decided to keep the dots.

[jedm]

Dots should be removed. There's too little space available to waste it.

[byndcivilization]

We need something to differentiate. Otherwise the text tends to flow together.

[jedm]

I suggest a color change to the Annotation Title, so that each new Annotation (or result in Search results) pops out.

(Highlighting the matched word in Search still useful too.)

On Sep 24, 2015, at 1:33 PM, Chris Perry notifications@github.com wrote:

We need something to differentiate. Otherwise the text tends to flow together.

— Reply to this email directly or view it on GitHub https://github.com/NRGI/resourcecontracts.org/issues/367#issuecomment-143043188.

[byndcivilization]

Looks good except there is some inconsistencies in page displays:

[SamCCSI]

I'm coming in late to this one but wanted to point out various errors in spelling in those annotations - they look as though they are auto-generated from OCR text that has not been MT'ed?

Eg. Heng YL'E (should be Heng Yue) Eg. IN7EXNA7IONAL (should be International)

Let me know if I should log as a separate issue

On Tue, Sep 29, 2015 at 2:56 PM, Chris Perry notifications@github.com wrote:

Looks good except there is some inconsistencies in page displays:

[image: unknown-5] https://cloud.githubusercontent.com/assets/2875950/10174926/45d6688c-66ba-11e5-9216-b5de356ba106.png

— Reply to this email directly or view it on GitHub https://github.com/NRGI/resourcecontracts.org/issues/367#issuecomment-144156083 .

SAM SZOKE-BURKE _Legal Researcher_Columbia Center on Sustainable Investment Columbia Law School - The Earth Institute, Columbia University Level 1, Warren Hall, 410 W116th St. New York, NY 10027 | (212) 854-2635 <212-854-2635> | (646) 630-4123 <646-630-4123> | s.burke@columbia.edu | www.ccsi.columbia.edu | @CCSI_Columbia https://twitter.com/CCSI_Columbia | @samszokeburke http://twitter.com/samszokeburke

[anderspeders]

@SamCCSI Thanks for catching. Can you log under #402?

[SamCCSI]

I'm thinking this might just have been a mock-up of annotations and that those typos are annotation text, not category text. If so, this is not an issue, as the errors are just copied from OCR that hasn't been MT'ed. @byndcivilization https://github.com/byndcivilization does that sound right?

On Tue, Sep 29, 2015 at 3:21 PM, Anders notifications@github.com wrote:

@SamCCSI https://github.com/SamCCSI Thanks for catching. Can you log under #402 https://github.com/NRGI/resourcecontracts.org/issues/402?

— Reply to this email directly or view it on GitHub https://github.com/NRGI/resourcecontracts.org/issues/367#issuecomment-144161651 .

SAM SZOKE-BURKE _Legal Researcher_Columbia Center on Sustainable Investment Columbia Law School - The Earth Institute, Columbia University Level 1, Warren Hall, 410 W116th St. New York, NY 10027 | (212) 854-2635 <212-854-2635> | (646) 630-4123 <646-630-4123> | s.burke@columbia.edu | www.ccsi.columbia.edu | @CCSI_Columbia https://twitter.com/CCSI_Columbia | @samszokeburke http://twitter.com/samszokeburke

[byndcivilization]

Annotations aren't OCRed or mturked. I think those search results are on contract text.

[anderspeders]

@SamCCSI yes, just looking again at these. The displayed image is search results in the OCR-test and thus these will have errors from time to time.

[SamCCSI]

Thanks, I'm in the process of noting all contracts that we will hide text view (because not yet MTed) so presumably there wont be search results for those contracts, and hence no chance of ugly pre-MT text?

cc @kaitlinccsi

On Tue, Sep 29, 2015 at 3:38 PM, Anders notifications@github.com wrote:

@SamCCSI https://github.com/SamCCSI yes, just looking again at these. The displayed image is search results in the OCR-test and thus these will have errors from time to time.

— Reply to this email directly or view it on GitHub https://github.com/NRGI/resourcecontracts.org/issues/367#issuecomment-144166529 .

SAM SZOKE-BURKE _Legal Researcher_Columbia Center on Sustainable Investment Columbia Law School - The Earth Institute, Columbia University Level 1, Warren Hall, 410 W116th St. New York, NY 10027 | (212) 854-2635 <212-854-2635> | (646) 630-4123 <646-630-4123> | s.burke@columbia.edu | www.ccsi.columbia.edu | @CCSI_Columbia https://twitter.com/CCSI_Columbia | @samszokeburke http://twitter.com/samszokeburke

[anjesh]

The search results will still appear, as we are hiding the text with "Show text" option only. We will fix the search result as well. Thanks for pointing it out.

[jedm]

Would be useful to know if the temporary fix involves omitting the raw text from searches on a per-contract basis or using some other method.

(And per a question to Anders a few weeks ago, would be good to have confirmation of what is being searched by default (i.e., Annotation text, Annotation titles, Metadata categories, Clusters) and with what weighting. Explanation of this highly useful but not urgent to launch at all.)

[anjesh]

Right now the search in the contract view page is to search the pdf-text only, and clicking will scroll to the particular page. I haven't had any discussion regarding searching annotations and metadata. cc @anderspeders

[jedm]

It would be a surprise if Annotation text is not being searched, since it is of especially high value because it's been created via expert review. Please do confirm at your convenience what is searched and what is not. (Also, by pdf-text do you mean OCR text or something else?)

Again, this is important to understand but not time-sensitive.

[anjesh]

By pdf-text, i mean OCR text. Lets discuss what else to expose in this search. Taking off Sprint 9.

[anjesh]

@anderspeders so we need to have search for annotation as well in the current search box

Currently it searches only OCR text (if available), now with annotations, they will be listed side-by-side differentiated with annotation and text icons besides the results.

[SamCCSI]

cc @kaitlinccsi

On Tue, Oct 13, 2015 at 1:50 AM, Anjesh notifications@github.com wrote:

@anderspeders https://github.com/anderspeders so we need to have search for annotation as well in the current search box

[image: image] https://cloud.githubusercontent.com/assets/120488/10446817/66501466-719e-11e5-8b42-2add672a253c.png

Currently it searches only OCR text (if available), now with annotations, they will be listed side-by-side differentiated with annotation and text icons besides the results.

— Reply to this email directly or view it on GitHub https://github.com/NRGI/resourcecontracts.org/issues/367#issuecomment-147613110 .

SAM SZOKE-BURKE _Legal Researcher_Columbia Center on Sustainable Investment Columbia Law School - The Earth Institute, Columbia University Level 1, Warren Hall, 410 W116th St. New York, NY 10027 | (212) 854-2635 <212-854-2635> | (646) 630-4123 <646-630-4123> | s.burke@columbia.edu | www.ccsi.columbia.edu | @CCSI_Columbia https://twitter.com/CCSI_Columbia | @samszokeburke http://twitter.com/samszokeburke

[KaitlinCCSI]

Can you provide a mockup of what this will look like? Unclear to me.

[anjesh]

Created new issue to enable searching for both annotation and text in the same left side box. https://github.com/NRGI/resourcecontracts.org/issues/761