Typically, functionality descriptions of the discrete electronic components are reported in datasheets. Annually, thousands of new components are designed and
added to the multitude of existing ones. So for this purpose, we implemented a new tool to detect the datasheet title (circuit category) and then extract relevant performance specs.
The datasheet scrubber automatically processes these documents and summarizes their performance specifications withing text or tables. Moreover, for each component, it recognizes the main functionality of each pin, e.g. power, ground, reset, etc. which is an essential step for automatically connecting components.
We have published a paper on Elsevier Array journal based on this project.