Printable page

Sophic


Sophic Intelligence

Sophic Intelligence (patent pending) is an advanced linguistic, text mining and visualization system integrated into the Biomax BioXM Knowledge Management System. Sophic Intelligence and BioXM use sophisticated query structures, advanced methods, processes and algorithms to find and map semantic and scientific relationships between key words or phrases found in large, complex, text files and related elements stored in structured databases. Sophic Intelligence provides users a unique two “Google-like” windows WIKI interface that allows the user to type, or cut and paste, a string of key words related to a specific area of interest in each of the query windows. Sophic Intelligence launches an “all against all” query (a parser on the fly) across all documents and databases to generate a “heatmap” visualization of the related semantic and scientific elements buried in disparate data silos.

These semantic relationships are then used to generate heatmaps that show specific color-coded signatures that are directly related to frequency and intensity of information often buried and overlooked in large, complex text files. These intuitive color coded heatmaps not only show users what is in a massive amount of text, but also clearly highlights what is not in documents. This type of fit-for-purpose analysis generates unique and significant “intelligence” value.

Sophic Intelligence is a multi-purpose system, configured and integrated into SCan-MarK Explorer to allow mining of all cancer-related Medline abstracts and the mining of FDA Drug Labels and DrugBank in the Pharmaceutical Drug Safety system.

  • Sophic Intelligence
  • Query Results
  • Pairs of Related Terms from Different Sources
  • Sentences View
  • Cancer Gene Index Evidence View

Sophic Intelligence allows the user to formulate a relationship based query by entering terms in two “Google-like” windows. Enter terms in window one and related terms in window two followed by a date range. Sophic Intelligence generates an “all against all” parser on the fly and launches the query across both structured and unstructured data sources.

Query results are displayed in heatmaps showing the frequency of the co-occurrence of related objects terms such as genes or NCI thesaurus. This highlights the “density” of information retrieved from across disparate sources and it shows “green fields,” where there is little or no information on a pair of terms. The numbers of co-occurrences in the table are linked to the source. Below the heatmap is a table with Pairs of Related Terms and a link to the source of the information.

The sources of information for a related pair, DKK1 and Breast carcinoma, are shown with evidence from text mining and the Cancer Gene Index project.

The sentences view presents the sentences that contain the co-occurrence of the terms from the query results. This view allows you to click on the link to get to the full abstract in Medline.

The Cancer Gene Index (CGI) evidence view presents the number of CGI relations between DKK1 and Breast Carcinoma from the query results. This view shows the evidence from the CGI project and allows to access full abstract from Medline.


© 2014 Sophic Alliance Inc. (508) 495-3801  - Webmaster: Richard Navin - Hosted by: Diversified Technologies