org names

[mindrones]

[lizgzil]

@mindrones again, in this case this is what GlassAI have said is the company name. So I guess I will include in an email to them.

There are a few other dodgy long names (which also involve dashes). So for now, I suggest we edit text from the names if

• total number of characters in the name is >35
• there is a ' -' or '- ' in the name
• remove everything including and after the ' -'

This would clean up:

• 'Peptone - The Protein Intelligence Company'
• 'lifesdna - wellness wellbeing healthcare data blockchain, AI powered search engine marketplace'
• 'MICRA - Midlands Innovation Commercialisation of Research Accelerator'
• 'S2T Unlocking Cyberspace - AI-Powered Investigations,WEBINT, OSINT, FUSION'
• 'BlackCurve - Price Management & Optimisation Software'
• 'Couvenda™ - the 24/7 concierge for your brand'
• 'Beyond Image - Intelligent Massage Retreat'
• 'Aimera - Interactive Marketing Channel'
• 'InScribe - ML Handwriting Tool for iPad'
• 'Inzura - Creators of the iApprove Digital Insurance App'
• 'EnrichApp - Simplifying Healthcare Collaboration'
• 'Food & Environment Research Agency -FERA'
• 'Barclays Accelerator, powered by Techstars - London']

Names with dashes in that wouldn't be cleaned are: ['Q-Equip', 'Q-Bot', 'Qubitl - Quantum Technologies', 'Planning-inc', 'pi-health', 'Lightfi - True Retrofit Solution', 'Kompli-Global', 'Mesh-AI', 'Miltech-AI UK Ltd', 'Uni-Soft Systems Ltd.', 'Test-Track-Monitor Limited', 'x-Sentience Limited', 'Work-books Ltd', 'XMED iQ - International Group', 'Vet-AI', 'Spectra-Scope UK', 'Sky-Futures', 'Smart-cot', 'Simply Regulation - FinregE', 'Social-i Networks Ltd', 'Blue Duck Education - Mangahigh.com', 'BOOKLUK.com - Bookluk Ltd', 'BOT-IOT', 'Connect-AI', 'AI-Detect', 'AI-PRIORI Ltd.', 'AI-Adam', 'A-lead', 'A-dapt', 'Ad-Lib Digital', 'BIAMI.io - Intelligent Automation', 'Aurora-AI', 'Aurora-ecs', 'Apricus Med-Tech UK Ltd', 'analytics-link', 'ArtiFit - AI Personal Trainer', 'ARG-tech', 'Hawk-Eye Innovations', 'G-Research', 'Hyper-Group', 'I-LAPS', 'iLoF - Intelligent Lab on Fiber', 'Digital-Dandelion', 'E-Nano', 'Earth-I', 'Farm-Hand Ltd', 'feather-ai', 'elementary-AI', 'eQuoo - The Emotional Fitness Game', 'STFC - Laboratories', 'Co-FundNI', 'London Co-Investment Fund']

From these arguably

• 'Lightfi - True Retrofit Solution'
• 'Blue Duck Education - Mangahigh.com'
• 'BOOKLUK.com - Bookluk Ltd'
• 'BIAMI.io - Intelligent Automation'

• 'eQuoo - The Emotional Fitness Game' should also be cleaned. But I guess with a rule based approach we can't get everything.

  So I wonder:

  • shall I implement the rule described above to get most of the problems?
  • shall I hard code the cleaning of those other examples or just leave them as they are?

  Note: these examples are from the full output (GTR, Crunchbase)- not just GlassAIs contributions.

[mindrones]

@lizgzil Thanks for this.

Would it make sense to remove the text after the "spaced dash" and capture it as a new field? That might help in the UI to manage "extras" for example shortening them based on the available screen estate or so.

[lizgzil]

@mindrones yes I can put it in a new field for those 13 or so organisations ("name_extras" say?). So, just to check we are on the same page, it'd look something like:

[{'description': '...',
  'location': {'lat': xxx, 'lon': xxx, 'postcode': 'xxx'},
  'name': 'lifesdna',
'name_extras': 'wellness wellbeing healthcare data blockchain, AI powered search engine marketplace'
  'place_id': 'xxx',
  'types': [x],
  'url': 'xxx'}]

[mindrones]

@lizgzil name_extras (or name_extra) would be ok to me 👍

I was thinking to apply this to all orgs, not necessarily those with name longer than 35 chars, so that e.g. eQuoo - The Emotional Fitness Game would be:

{
  ...,
  "name": "eQuoo"
  "name_extra": "The Emotional Fitness Game",
  ...,
}

[lizgzil]

ah I see - that works!