Tool to visualize boolean logic/ search queries
Hi there, is there a tool to visualize search queries? The Most Regards You’ve Ever Been Sent 😊 Researcher / Content Creator & Curator/ Writer (Psychology, Career Choice & Development) Hashem ElAssad Follow me on LinkedIn<https://www.linkedin.com/in/hashemelassad/>
Yes. DiscoverText is a web-based tool, free for academics who sign up with a valid academic email, with visual search results using highlights as one of five key pillars of text analytics. This is year 17 of my team engineering a system for conducting scientific, mixed-methods research using human annotators to crowdsource and measure human reliability and accuracy. Boolean search is a fundamental building block of what we do. Search, filtering, deduplication, clustering, human annotation, measuring inter-rater reliability, adjudication of disagreements, ranking of annotators over time, and machine-learning (ML) are integrated parts of a platform where the visualization of search is one of the greatest research accelerators I have ever used. I will often set the list view to 500 or 1,000 results per page and just scroll through the highlights for a deep immersion experience. We also enable labeling off the list view, so I combine Boolean search results with ML predictions to produce extremely clean lists where outliers are more easily identified. Finally, I use a browser add-on called "HighlightThis" to create another layer of highlights for particular subsets of discursive elements I am studying. Here are some simple Boolean queries visualized in a short slide deck: https://tinyurl.com/visualsearchresults The dataset in this case is an interesting one. Most of these examples in the slides are from November 6-7, 2020. The original Boolean collection query was "Dominion AND (voting OR systems OR system)". There is an as yet untold story about what happened in the first hours and days of the Dominion Voting Systems uproar that cost Fox News $750 million. I am currently looking for a collaborator or team to work with to write that up. It is probably the most interesting 400,000 Tweets I have ever encountered. For more information on our approach to "humans and machines learning together" there are a few short introduction videos. https://tinyurl.com/DTBasics If you would rather use built tools than spend your time building them, please book a meeting: https://calendly.com/discovertext ~Stu On Sun, May 19, 2024 at 8:35 AM Hashem Elassad via Air-L < air-l@listserv.aoir.org> wrote:
Hi there, is there a tool to visualize search queries?
The Most Regards You’ve Ever Been Sent 😊
Researcher / Content Creator & Curator/ Writer (Psychology, Career Choice & Development)
Hashem ElAssad
Follow me on LinkedIn<https://www.linkedin.com/in/hashemelassad/>
_______________________________________________ The Air-L@listserv.aoir.org mailing list is provided by the Association of Internet Researchers http://aoir.org Subscribe, change options or unsubscribe at: http://listserv.aoir.org/listinfo.cgi/air-l-aoir.org
Join the Association of Internet Researchers: http://www.aoir.org/
-- Dr. Stuart W. Shulman Founder and CEO, Texifter Editor Emeritus, *Journal of Information Technology & Politics* ResearchGate Profile <https://www.researchgate.net/profile/Stuart-Shulman>
Thanks Stu, This email was as much a potential influence akin to what Grimmer and Stewart's article had on you. Amazing presentation on humans and machines learning together spiced with your own learning journey. 👏👏 best, rb On Mon, May 20, 2024, 4:56 PM Shulman, Stu via Air-L < air-l@listserv.aoir.org> wrote:
Yes. DiscoverText is a web-based tool, free for academics who sign up with a valid academic email, with visual search results using highlights as one of five key pillars of text analytics. This is year 17 of my team engineering a system for conducting scientific, mixed-methods research using human annotators to crowdsource and measure human reliability and accuracy. Boolean search is a fundamental building block of what we do. Search, filtering, deduplication, clustering, human annotation, measuring inter-rater reliability, adjudication of disagreements, ranking of annotators over time, and machine-learning (ML) are integrated parts of a platform where the visualization of search is one of the greatest research accelerators I have ever used. I will often set the list view to 500 or 1,000 results per page and just scroll through the highlights for a deep immersion experience. We also enable labeling off the list view, so I combine Boolean search results with ML predictions to produce extremely clean lists where outliers are more easily identified. Finally, I use a browser add-on called "HighlightThis" to create another layer of highlights for particular subsets of discursive elements I am studying.
Here are some simple Boolean queries visualized in a short slide deck: https://tinyurl.com/visualsearchresults
The dataset in this case is an interesting one. Most of these examples in the slides are from November 6-7, 2020. The original Boolean collection query was "Dominion AND (voting OR systems OR system)". There is an as yet untold story about what happened in the first hours and days of the Dominion Voting Systems uproar that cost Fox News $750 million. I am currently looking for a collaborator or team to work with to write that up. It is probably the most interesting 400,000 Tweets I have ever encountered.
For more information on our approach to "humans and machines learning together" there are a few short introduction videos. https://tinyurl.com/DTBasics
If you would rather use built tools than spend your time building them, please book a meeting: https://calendly.com/discovertext
~Stu
On Sun, May 19, 2024 at 8:35 AM Hashem Elassad via Air-L < air-l@listserv.aoir.org> wrote:
Hi there, is there a tool to visualize search queries?
The Most Regards You’ve Ever Been Sent 😊
Researcher / Content Creator & Curator/ Writer (Psychology, Career Choice & Development)
Hashem ElAssad
Follow me on LinkedIn<https://www.linkedin.com/in/hashemelassad/>
_______________________________________________ The Air-L@listserv.aoir.org mailing list is provided by the Association of Internet Researchers http://aoir.org Subscribe, change options or unsubscribe at: http://listserv.aoir.org/listinfo.cgi/air-l-aoir.org
Join the Association of Internet Researchers: http://www.aoir.org/
-- Dr. Stuart W. Shulman Founder and CEO, Texifter Editor Emeritus, *Journal of Information Technology & Politics* ResearchGate Profile <https://www.researchgate.net/profile/Stuart-Shulman> _______________________________________________ The Air-L@listserv.aoir.org mailing list is provided by the Association of Internet Researchers http://aoir.org Subscribe, change options or unsubscribe at: http://listserv.aoir.org/listinfo.cgi/air-l-aoir.org
Join the Association of Internet Researchers: http://www.aoir.org/
participants (3)
-
Hashem Elassad -
Rakesh Biswas -
Shulman, Stu