> >Well, the team has achieved about 97% accuracy for the specific parts of the SDS they were testing, such as the boiling points, flash point, Hazard Statements, etc.
>
Thanks for the update on this. I am correct in assuming that part of this impressive accuracy rate can be attributed to the structured way that the data on the SDS's is organized? Is it possible to achieve this level of accuracy for more free form text on SDSs such as descriptions of reactivity, toxicity data or waste disposal instructions. (Although I suppose that waste disposal instructions are likely to be very boiler plate as well on SDSs).
I also wonder what sources the chemist who reviews the results will use to assess the data extraction results. Reference texts, other SDSs, other? It appears that many times a piece of data is repeated from one document to another without attribution to its source, which can lead to a garbage in, garbage out situation‰?|
As I mentioned in my national meeting presentation yesterday, I am concerned that the transfer of risk information between organizations as development of a chemical is outsourced and the use of a chemical is globalized through technology transfer is not well-established. So thanks for sharing your experience with the broader community. I suspect this will become an increasingly important issue over time.
- Ralph
Ralph Stuart, CIH, CCHO
ralph**At_Symbol_Here**rstuartcih.org
---
For more information about the DCHAS-L e-mail list, contact the Divisional membership chair at membership**At_Symbol_Here**dchas.org
Follow us on Twitter **At_Symbol_Here**acsdchas
Previous post | Top of Page | Next post