CV is normally an extremely rich document, and any CV parsing application must be able to extract and normalize a maximum of this information. H_Know
is able to extract the following types of data:
- Education – for each education event it extracts:
- The start and end date
- The type of education (formalized according to several national and international taxonomies)
- The institution providing the qualification
- The geographical location
- Experience – for each experience mentioned by the candidate it extracts:
- The start and end date
- A classification of the experience into a predefined taxonomy
- The set of concepts characterizing such an experience
- The company/institution
- The place
- Skills – for each skill declared by the candidate, H_Know extracts:
- The primary skill, classified according to some of the major skill ontologies (LinkedIn, Janzz ISSCO)
- The secondary skills
- Additional Information:
- All contact information: email, first name, family name, address, telephone, etc.
- Languages spoken, classified according to a standardized degree of knowledge
- Any kind of personalized information the customer might wish to retrieve
Ho2S was the coordinator of what was probably the biggest international research effort on CV parsing : the SAUGE project (Semantic Analysis for Unrestricted Generalized Employment – see the Research page for more details). The goal of the project was to deliver the most effective CV Parsing technology by coupling statistical dependency parsing, provided by the University of Oslo and Ho2S, with the most fine-grained and market-validated ontology in the job domain, provided by JANZZ. By bridging the world of linguistic parsing with the semantic layer provided by the Linked Open Data paradigm, the project created a valuable, and potentially infinite, network of knowledge around any CV, in most European languages and in virtually all domains of human activity. Ho2S integrated this revolutionary technology into its products as of the second half of 2014.
t is perfectly clear to technical people why CV parsing represents a crucial technology in the cycle of handling human resources. This is sometimes less clear for those that sit on the margins of the HRM business.
In simple terms, parsing a CV enables you to answer questions for which no response was previously available. For instance, in the following scenarios:
- I want to select a person with skills in Java programming, who has worked as a requirements engineer in a multinational company and who has a PhD obtained no more than three years ago.
- I am a public decider and I would like to know the degree of employment of people who graduated from a particular school in a specific European state.
- I am the head of HR and I need to detect, in my organization, a high-level profile with very good knowledge of German, skills in solar energy and possibly with experience studying abroad.
- I am operating a competitive intelligence program and I want to know, out of a statistically relevant sample, how many people were hired as systems administrators by my competitor and are no longer in that company.