We have been comparing ChatGPT responses using statistical similarity v knowledge representations in the automated text selection process. There are token limits with Gen AI tools like ChatGPT meaning the selection of the subset of text from a search index is critical. Whilst statistical similarity of sentences or chunks of sentences that match the initial... Continue Reading →

Part of the knowledge representation used by GeoClassifier(R) to automatically classify documents by Document Type. There are too many types, clues and synonyms to display as labels so they grouped by discipline in this visualisation. This is a poly hierarchy so the same document type can exist under two disciplines. For example 'HSE well notification'... Continue Reading →

Information visualisation of the reference knowledge graph used by the GeoClassifier(R) algorithm to automatically classify subsurface documents to detailed document type level. Whilst machine learning, generative AI and NLP techniques are also used by GeoClassifier(R), this large reference knowledge graph helps the algorithm understand the domain. More: contact@infosciencetechnologies.com

PETRONAS and Infoscience Technologies exhibiting their partnership at Petronas Digital Day May 29th to May 30th 2023. Eric Andersen (Right) Petronas Head of Geoscience Solutions and Chief Geoscientist Exploration and Development with Petronas colleagues. The event highlighted transformative digital solutions and explored business ideas for this year’s theme - Digital in Motion, which aims to... Continue Reading →

https://player.vimeo.com/video/827345849?h=9bbe471e3f&badge=0&autopause=0&player_id=0&app_id=58479

Applying ChatGPT to subsurface and wells content. We have been applying OpenAI’s ChatGPT to the graph data structures produced by the OpportunityFinder(R) and GeoClassifier(R) NLP algorithms after they have processed unstructured data (reports, presentations, literature etc) at a sentence level. This technique called prompt engineering, allows you to apply Large Language Model structure but on... Continue Reading →

Infoscience in the news today from the Geological Society of London. https://geoscientist.online/section/news/

Pleased to announce Infoscience Technologies Ltd has received the King’s Award for Enterprise Innovation. The award was conferred today by His Majesty King Charles III to Infoscience Technologies for outstanding achievement in innovation after recommendation by the Prime Minister. The award is the most prestigious for a UK business. Infoscience Technologies' patented Artificial Intelligence algorithms... Continue Reading →

Image: A small public document set 'spatialized' using Infoscience's GeoClassifier algorithm, each dot is a document. GeoClassifier uses novel techniques to disambiguate - automatically assigning latitude and longitude points for any document. Ready to integrate in a company's GIS system. #naturallanguageprocessing #machinelearning More: contact@infosciencetechnologies.com

GeoClassifier v4 uses Natural Language Processing (NLP) and Machine Learning to classify and position documents on a map. The spatialisation includes many unique features around disambiguation and detection of the most specific latitude and longitude. Significantly improves on many existing and rather basic geotagging options in use today. #geotagging #spatialisingdocuments #georeferencing More at: contact@infosciencetechnologies.com