SourceCon’s Semantic Search Series

How’s that for an alliteration?

Back in 2007, Dave Copps introduced us to PureDiscovery at the very first SourceCon conference, and in doing so he opened the door to semantic search for many of us in the sourcing world. Semantic search is not a new concept, but it’s one that alludes many of us because it’s just one of those things that has a bit of a fuzzy definition.

Semantics, by its purest definition, is the study of the meaning of something, usually a language. This is not to be confused with syntax, which is the grammatical arrangement of words in sentences. Semantics is more about the words chosen to share a message, where syntax is more about how those words are arranged within the message. The two are interconnected, but they will take you in different directions when you search.

The term semantic web refers to how our computers and internet technologies take a look at the information out there on the Web and make sense of the meaning of it all. This specifically includes social networks because a lot of what’s contained in social networks are thought streams of individual people. The original idea of a semantic web is credited to World Wide Web Consortium (W3C) director Tim Berners-Lee, when he shared the following thought back in 1999 in his book “Weaving The Web“:

“I have a dream for the Web [in which computers] become capable of analyzing all the data on the Web – the content, links, and transactions between people and computers. A ‘Semantic Web’, which should make this possible, has yet to emerge, but when it does, the day-to-day mechanisms of trade, bureaucracy and our daily lives will be handled by machines talking to machines. The ‘intelligent agents’ people have touted for ages will finally materialize.”

The purpose of semantic search technology is to understand the intent of the searcher. What semantic search engines do is look at words surrounding words and attempt to “disambiguate” them. Lots of words have multiple meanings, and semantic search technology takes a look at the context in which the words appear and attempts to clarify what it believes the intent of the searcher is. The hope of those who create semantic technologies is that they will learn search patterns over time and be able to better predict what the human searcher is trying to find. (see Google’s most recent Instant Search product release)

Article Continues Below

The technology behind semantic search engines is quite complicated. But we thought it might be nice to examine some of these search engines a little more closely to see how they can be useful to us in our daily sourcing functions. So in the coming weeks, we will be reviewing several semantic search tools. This will include (but is not limited to):

If you’d be interested in conducting a review of a semantic search tool that can be posted as part of this series, please let us know as well. We’d love to include as many tools as possible to give sourcers some good information on semantic technologies and how they will be helpful to us as we search for leads and candidates.

Happy Sourcing!

Amybeth Quinn began her career in sourcing working within the agency world as an Internet Researcher. Since 2002, she has worked in both agency and corporate sourcing and recruiting roles as both individual contributor and manager, and also served previously as the editor of The Fordyce Letter, and, with ERE Media. These days she's working on some super cool market intelligence and data analytics projects. You can connect with her on Twitter at @researchgoddess.


3 Comments on “SourceCon’s Semantic Search Series

  1. Thanks, Glenn! We’re actually going to include TalentSpring along with some other specific tools to the recruiting world in this series. It is always good to be able to reference and learn from the research done by other industry colleagues for things like this!

  2. I felt a disturbance in the force.

    Did someone say “semantic search?” 🙂

    Semantic web search (e.g., Kngine, Hakia, etc.) and semantic search applied specifically to human capital data are two entirely different animals.

    I *love* technology, but I always approach vendor semantic search solutions with a healthy dose of rational skepticism, and the current state of semantic search applications for sourcing/recruiting have their limitations and it is important to understand them as well as how they (or how they don’t) achieve their claims.

    I don’t have a semantic search product to sell, so I often feel I am the only one talking about “manual” semantic search, which only requires a search engine that supports configurable proximity and term weighting, which allows the targeting of meaning at the sentence (and thus actual responsibility) level, which is an order above the word, phrase, and statistical semantic matching that ALL automated (aka, do the thinking for you) solutions on the market today target.

    With regard to machine learning, semantic clustering and semantic query clouds based on statistical pattern recognition, it is critical to always remember that related terms are not necessarily relevant – that’s a key distinction few people discuss.

    For those who are interested, here is a link to my SourceCon keynote on resume sourcing and matching: artificial intelligence vs. human cognition.

    I’ve also written several articles on the subject of semantic search for sourcing:

    BTW – I’d love to have the opportunity to work on developing the next generation of semantic search technology. Any takers?


Leave a Comment

Your email address will not be published. Required fields are marked *