Connor Forbes

[Connor Forbes @connorf25]

@sama Why not release open source if it is being retired? I thought the company is called OpenAI

[Pavel Zhelnov, MD @drzhelnov]

Congrats Connor and everyone at @Institute4EBH on publishing a paper on Deduplicator https://t.co/4vs3073LV2 (I’ve been looking forward to it!). Haven’t read yet, but an F1 score of 0.9645 looks great! Glad to finally have a full-fledged paper to cite😊 https://t.co/x3hSugHBFo

[Connor Forbes @connorf25]

@I_AM_AbouSetta @StephanieRothSR @EQUATORNetwork Haven't gotten a chance to work on a mobile interface at this stage, am currently working on improving the desktop version. Not to say there may someday be a mobile version, if I get some time to slot development in between my student work.

[Institute for Evidence-Based Healthcare @Institute4EBH]

📣Automation tools to improve the speed of conducting #systematicreviews. Cost $55AUD. This ONLINE workshop will provide an overview & live demo of the SRA tools. 2 Feb: 8-9:30pm (UTC+10, Europe suitable). More info & other dates: https://t.co/8NbaUJnCYR @2weekSR @PaulGlasziou

[Connor Forbes @connorf25]

@DrIanKellar @drzhelnov @2weekSR Appreciate the kind words Ian 😀

[Dr Ian Kellar @DrIanKellar]

Put my MSc Systematic Review module on YouTube. Covers some history of critical appraisal, PICO, PRESS, Polyglot, SR Accelerator, Endnote, using a MECIR compliant data extraction tool, RoB2 and the RoB2 macro, PRISMA2020, MMAT and SWiM. Hope it's useful - https://t.co/cpiD4FS5ws

[Connor Forbes @connorf25]

@nealhaddaway @Institute4EBH @TheLensOrg Allows uploading an xml, ris or nbib file and will automatically read pmid and doi from all input references and perform a union on the results (i.e. will search pmid and doi at same time)

[Institute for Evidence-Based Healthcare @Institute4EBH]

@Institute4EBH is pleased to announce the release of a new tool, SpiderCite! Spidercite finds forwards & backwards citations from an input library, through collaboration with @TheLensOrg. Find Spidercite at: https://t.co/qv7DPmZPY0
Inspired by @nealhaddaway's tool citationchaser

[Dr Danielle Pollock @Daniellep89]

Duplication is a common issue in reviews as we search across databases. The deduplicator tool could make this process easier. Thanks Connor #EE21

[Connor Forbes @connorf25]

@tomroper @Institute4EBH Hi Tom,
Thanks for spotting the line translation issue, have fixed it up in the latest version of Polyglot and it should now be working fine

[2weekSR @2weekSR]

Deduplicator now has 2 algorithms:
1. Cautious: Optimal for large libs, will identify few non-dups as dups (false +ves), 99.9% precision
2. Thorough: Optimal for small libs, will miss v. few dups; may need a check to ensure no non-dups wrongly classified
https://t.co/goqdUqoPMm

[Connor Forbes @connorf25]

@DrIanKellar @2weekSR Ahahah good question, the original Deduplicator actually predates Screenatron and Disputatron, although I would not be at all opposed with the name Dedupatron 😉

[Connor Forbes @connorf25]

@DrIanKellar @2weekSR @Institute4EBH @DrIanKellar Have just pushed an update to Polyglot, should now handle the cases you have outlined below. Let me know if you find any more cases 😀

[Connor Forbes @connorf25]

@DrIanKellar @Institute4EBH @2weekSR Doing some testing now and "(TI=((control? OR prospectiv?) NEAR/3 (trial? OR method? OR stud?)) OR AB=((control? OR prospectiv?) NEAR/3 (trial? OR method? OR stud?)))" also seems to work (should be fetching 421,819 results right?)

[Connor Forbes @connorf25]

@DrIanKellar @Institute4EBH @2weekSR Oh wow, don't even want to imagine the complexity of that one. Im guessing 24 individual NEAR/3 comparisons are going to be needed (12 for TI and 12 for AB) 😢 Really suprising that WOS haven't implemented this expansion in their engine

[Connor Forbes @connorf25]

@DrIanKellar @Institute4EBH @2weekSR Nonetheless, I will look into potential integrations with Polyglot, one potential solution is to perform this expansion for WOS (as above), or to simply replace NEAR/3 with AND in this case. Will look into how complicated coding the 1st solution will be

[Connor Forbes @connorf25]

@DrIanKellar @Institute4EBH @2weekSR (TI=(muscle NEAR/3 resist) or TI=(muscles NEAR/3 resist) or TI=(muscle NEAR/3 resists) or TI=(muscles NEAR/3 resists)) or (AB=(muscle NEAR/3 resist) or AB=(muscles NEAR/3 resist) or AB=(muscle NEAR/3 resists) or AB=(muscles NEAR/3 resists))

[2weekSR @2weekSR]

In honour of @WorldEBHCDay, the #2weekSR team will unveil & demo our cool new tool – the MethodsWizard. Assists with writeup of the methods section of systematic review protocol. Featuring: @PaulGlasziou, yours truly & @connorf25. Registration & details:
https://t.co/N7qAHGII4C