Flare’s approach to search extends the basic concept of scanning or crawling sources and making everything visible in one environment. This simplistic approach may work across small datasets or during pilots which expose ‘clean’ content, but is generally inadequate when scaled up to the enterprise level.
Users find that too many search results are returned, similar items are listed multiple times, or duplicates contaminate results.
Flare’s approach leverages the concept of tagging across multiple classes or facets to facilitate improved recall, precision and ranking. Flare considers search to be the visualization of content, and the sharing as a continuum.
When bringing together multiple sources of information, each should be assessed for the effort required to catalog it, and the value it might bring to the organisation. Flare’s approach considers the needs of both the end user and the Information and Data Management (IM) community.
Below is a simplistic picture of what Search and good IM might look like:
RSS news feeds, or pages scraped from the web may well be auto-tagged and made visible, but what rules should be available to ensure that content of this nature is ranked lower than useful internal content?
When it comes to new digital content created by company professionals, shouldn’t the approach be different? What tools exist to help high-grade and add clean, accurate metadata to reports, some of which may have taken months to write? What if a version of a document was presented at a key meeting or submitted to a government to meet regulatory requirements? How might that be high-graded?
A more sophisticated view of what search and good IM might look like:
Notice that this picture does not consider the technology alone, but cultural aspects such as governance, and process and transactional information.
Flare provides several modules to deal with these challenges, which collectively contribute to an overall improved search experience. By considering not only what information should be made visible within a search engine, but also providing different methods to ingest that information, we can provide the end-user with a more intuitive, efficient and robust search experience.