Avature announced Thursday the beta release of its semantic search module. The first version will be tested by a small group of customers beginning in October and will then be made available to all customers as an add-on to their existing ATS solution in the first quarter of next year.
The semantic search module currently supports English language searches, but will be updated to include German language searches before the end of the year, and French language searches in the first quarter of 2019.
The module will be available to all ATS customers with a completely revamped search UI that supports both semantic and standard Boolean search. Avature’s federated search module, WebSources, which searches third-party systems from within the Avature CRM solution, will add semantic search features later in the year.
The module, an important component of Avature’s evolving AI strategy, is primarily designed to improve matching of candidates to jobs by using their resume content to expand candidate search criteria based on semantic correlations.
Semantic Search will join Saccade and Avature´s “Find Similar” features to form the cornerstone of Avature’s learning-based AI strategy. Saccade, Avature’s NLP parser, has already learned to parse resumes in over 20 languages, including Chinese and Japanese, while “Find Similar” is designed to find patterns in large sets of similar documents and use them as search parameters.
Central to this strategy is the concept that machine learning should be embedded in the platform so that massive amounts of data do not need to be exported to a separate system for AI processing.
“Unlike contracts or statements of work, resumes are difficult document types to search because of the abundant use of acronyms that mean different things in different industries, the frequent use of similar words with different meanings, key-word stuffing, and related concepts such as ‘manager of corporate development’ and ‘strategic alliance manager,'” said Nicolás Bader, Avature’s product marketing analyst for AI. “The key for Avature is to automatically compensate for these issues while reducing search complexity and improving the quality of search results.”
“Our recruiting customers want to improve the services they provide to their hiring managers,” said Dimitri Boylan, CEO of Avature. “With semantic search, recruiters can make expert level associations without any prior industry specific knowledge – selecting better candidates sooner, and improving the experience of the high-quality candidate. We think this will be particularly useful where recruiters need to manage job assignments outside their own area of expertise, and for areas where deep domain expertise is essential, such as pharmaceutical, medical technology, and health care recruiting.”
Avature will continue to look at how neural network technology and machine learning can leverage the knowledge that is accumulated inside an enterprise system and how this information can be combined with external information to improve user productivity and support higher levels of automation.
The module will be demonstrated at the upcoming EU Avature Conference in London on November 15.