Bitvore has added sentiment scoring for municipal obligors, allowing users to glean leading indicators of performance from unstructured data sources in real time.
Muni portfolio managers, credit risk underwriters and analysts can now use Bitvore obligor and CUSIP sentiment scores to trend and identify leading indicators of opportunity and risk without the efforts associated with manually sifting through news and other unstructured data sources. In addition to the obligor sentiment scoring, Bitvore also said it would include in its muni email alerts article-level sentiment scores.
“Obligor sentiment scoring has been one of the most widely requested enhancements by customers of our Bitvore for Munis product,” said Jeff Drake, CEO of Bitvore. “We’re pleased to announce continuously updated obligor sentiment scoring, allowing immediate identification of emerging opportunity and risk.”
Bitvore derives obligor sentiment by first using sentence by sentence analysis of ingested records that are determined to be business-related and about a specific obligor. Semantic orientation, linear sequencing and semi-supervised, deep learning are then used to derive sentiment at the sentence level.
The sentence level sentiment values are then rolled up to an overall aspect-oriented sentiment for each article. Bitvore then derives an average sentiment score in the Bitvore for Munis UI for a selected obligor or CUSIP and time period.
