Content classification

Category-Connect is MDPI’s tool for sorting scientific papers into research areas, assigning categories to manuscripts consistently instead of ad hoc.

Molecular Connections built a deep indexing service for a major chemical information provider: indexing key concepts, substances, reactions and structures across disciplines, including content from non-English journals. The multilingual pipeline handles Chinese, Japanese and Russian journals with automation,…

Molecular Connections built a high-throughput Keyword Recommendation Service for IFIS, the not-for-profit food and health sciences publisher. Deep learning models — semantic similarity, contextual vectorisation, bi-directional transformers — analyse the Food Science and Technology Abstracts database and suggest…

Molecular Connections built an alert system that sends society members topic-specific notifications, using machine learning and semantic enrichment to keep alerts relevant instead of flooding inboxes.

A lightweight tool from the STM Association that matches publisher content to the UN Sustainable Development Goals. It runs text against two independently developed sets of Boolean queries from bibliometrics companies, built on specific SDG targets rather than…

SDGenie (beta) is a free experimental tool from CABI that matches research text, such as abstracts, to the UN Sustainable Development Goals at the level of the 169 individual targets rather than the 17 headline goals. Released as…

Molecular Connections worked with a physical sciences publishing society (AIPP) on content discoverability: building a rich ontology and semantically enriching millions of backfile articles using semantic fingerprinting and content classification.

Elsevier is trialling ChatGPT in SSRN’s classification process, assigning research papers to topic areas within SSRN’s multi-level taxonomy. Runs on Microsoft Azure OpenAI, with Elsevier building its own production pipeline and testing incrementally for accuracy and rule adherence.