JANZZon! The unique, multilingual and most
comprehensive ontology. Worldwide.
Via JANZZjobsAPI, you can use the SaaS solution JANZZon! for your applications, the largest multilingual encyclopedic knowledge database in the area of occupation data (in particular, jobs, job classifications, hard and soft skills, training/qualifications, etc.) currently available. Both data-driven and expert consultation taxonomies are integrated into JANZZon! Amongst others, it covers O*Net, ESCO, DISCO II and UK skills taxonomy from Nesta. The occupation data is updated on a regular basis. JANZZon! follows the OECD Principles on Artificial Intelligence (AI).
JANZZon! is, for example, ideal for applications in the following areas:
- Job and training information systems
- (Job/skills) matching engines
- CV parsers and CV databases
- ATSs, CRM, ERP and other applications in the area of personnel recruitment
- Job portals, including social networks, career pages and aggregators
- Applications in the area of government labor market and placement measures
- Statistical analysis and modelling tools
- Full-text analysis and search tools
Thanks to its already extremely large number of poly-directional concepts, JANZZon! is able to generate unique conceptual relationships and facilitates previously unknown intelligent possibilities in the following areas:
- Data enhancement and indexing, including context and (syntactical and/or content based) request completion for full-text searches and CV parsers
- Multilingual semantic skills and job matching
- Automatic allocations of and between different classification systems
- Use of complex and unstructured data in the areas of gap analysis and benchmarking
- Various evaluations and applications in the areas of statistics, data modelling and big data
Finally turn your big (occupation) data into smart data with JANZZon!
Selected key figures for JANZZon!
|Number of hours invested in development to date||>240,000|
|Available languages (in progress or completed)||9 (2019)|
German (master language)
40 (in 2020) including
|Number of conceptual graphs||>2.62 million|
|Number of occupation/profession concepts||>140,000|
|Number of capability/skill concepts||>1,000,000|
|Number of specialization concepts||>57,000|
|Number of function concepts||>24,000|
|Number of soft skills/social skills concepts||>157,500|
|Number of training/qualifications||>422,000|
|Number of nodes, relations and relationships||>350 million|
|Number of terms (all languages)||>282 million|
|Standard classification systems (in progress or completed)||80|
|Number of occupation classes||7|
Private knowledge graphs in connection with JANZZon!
Within JANZZon!, there are various options for covering your very individual needs and requirements in a simple, cost-effective and sustainable manner. For example, JANZZon! also allows for specific private ontologies / knowledge graphs to be configured and integrated. Terms that are important to you can also be incorporated, regularly updated and maintained (e.g. organization-specific functions, product names and versions, abbreviations, certifications and diplomas, special thesauri).
Furthermore, it is possible to influence the use, prioritization and context of specific concepts. For instance, it is possible to use country-, company- or language-specific descriptions of terms, while you can also define special spellings or preferred terms in your applications as default settings. These may include job titles like «joiner» or, depending on the area, «carpenter».
Alongside the standardized occupation classes (OCs 1 to 7) and their weightings and relations available in JANZZon!, for example, different solutions are also possible in accordance with your requirements.
Build or buy?
Benefits of and reasons for using JANZZon! as a cloud solution.
A recurring and important question in the area of taxonomies and knowledge graphs is whether it is better to develop and maintain your own ontology or whether it makes sense to purchase one at some point or use one. This question is not only relevant to those companies which have already attempted to develop their own knowledge graphs, but which, due to a shortage of specialist knowledge, are no longer able to develop it further or, for example, are unable to maintain it as a result of insufficient resources and/or financial means. In recent years, hundreds of thousands of often very expensive digital graveyards with varying scopes have emerged in the area of occupation data. In most cases, ontologies are already outdated on the day of purchase and no longer up to date. This is particularly true if you are looking to cover extremely heterogeneous and dynamic fields of knowledge such as occupation data. Most solutions are not even suitable for modern applications such as skills and/or job matching.
In almost all cases, the facts thus invariably argue in favor of a solution such as JANZZon! JANZZon! provides you with a secure, sustainable and cost-efficient cloud solution that nevertheless allows you to use almost all options for the desired applications and areas of use. You also benefit from the wisdom of crowds: from the many other users of JANZZon! in different languages, sectors and specialist areas as well as from the ongoing enhancements and updates they make. Additionally, the use of a unique combination of the Mechanical Turk method and current machine-learning approaches enables us to continually enrich our ontology with updates and upgrades. All this without having to make available the, in most cases, extensive resources yourself and within the framework of clearly structured price plans tailored to the volume and scope of your usage.
Below you can find typical application examples of our SaaS solution JANZZon! Of course, other functions and applications that have not yet been implemented are possible. We would be happy to check new ideas and areas of use with you.
JANZZon! – application example 1: web portal for career videos
Improving the (multilingual) search and matching functions of a web portal for career videos
A web portal with video profiles of professionals in various fields, intended mostly for young users who want to inform themselves in greater detail about a specific profession or wider career opportunities, required improvements of the search and matching functions, above all in the fields of data enhancement, synonyms and related search terms/occupations/concepts. Multilingual functions were previously not possible.
Improvements made possible by JANZZ.technology
A simple link via JANZZrestAPI has realized a sustained improvement in the search and matching functions, for example through an extended glossary and by supplementing the required occupation data and terminology with synonyms, related occupations and professions as well as translations of the terminology drawn from the comprehensive ontology of JANZZon! in real time. Eight languages are currently available for JANZZon! Over the next three years, approximately 30 additional languages are to be added, including Chinese, Arabic and Russian.
Solution and methodology
- Detailed discussions with the project managers allowed for a clear conceptualization of the process, and areas with the greatest potential for improvement could easily be identified
- The data request protocol for JANZZrestAPI was drawn up and implemented within a very short period
Accomplishment and results for the client
Substantially improved (semantic) search and matching results in terms of quality and relevance, including in the context of related terms and concepts, which could not have been achieved with conventional full-text functions and matching engines. Stable and inexpensive link via JANZZrestAPI for real-time enquiries. This has also proven effective in terms of the performance of large quantities of database enquiries. In addition, the link to JANZZon! ensures that clients receive the very latest data at all times, thanks to the systematic maintenance and continuous expansion provided by JANZZ.technology. Possible future JANZZon! applications could concern the improved semantic and multilingual matching of existing video tags.
JANZZon! – application example 2: semantic search
Improved semantic search queries because of profound background knowledge
In contrast to conventional full-text searches, the semantic search with the help of JANZZon! uses extensive contextual and background knowledge in order to identify the right datasets and to filter out datasets that most conventional search engines would wrongly list in the search results.
The extensive contextual knowledge provided by JANZZon!, for example, allows the search to relate «knowledge of investments» with «knowledge of asset management» and also
«Microsoft Word» with «word processing» or «office application software». In addition, search queries also include synonyms and, if desired, designations and terms from other languages. This means that in addition to «CEO», terms such as «Geschäftsführer/-in», «Geschäftsleiter/-in» or «Managing Director» appear, but no mismatches such as «Assistant to the CEO» or «Secretary to the Managing Director».
The results and the precision of semantic search and matching processes are dependent upon the scope and depth as well as the quality and comprehensiveness of the applied contextual and background knowledge and upon the deployed ontology.
An extensive and multi-layered ontology such as JANZZon! also covers above all «real everyday language use» of users in the various applications, industries and languages. Without this widest-possible coverage, which sometimes also includes highly «creative user terminology», an ontology cannot be successful in the long term. In recent years, for example, the term «Employee for Copies and Archive» has reappeared as «Executive Document Manager», or «Cleaners» have been renamed «Facility Managers» – to cite just two curious developments in the field of job titles.
Moreover, many occupations, even very traditional ones, have been given more up-to-date designations in recent years. However, these are often peculiar titles that are rarely used in other contexts. For this reason, common and conventional terms such as «Butcher» or «Nurse» need to continue to be recognized, and not merely new titles such as «Meat Processing Professional» or «Healthcare Professional». Because JANZZon! is also fed to a large extent by the thousands of entries made by real users and based on sources such as real job offers in all industries, languages and regions or CVs of job seekers, the highest-possible degree of topicality and practicality in terms of search and matching processes is guaranteed.
However, as even extremely extensive ontologies such as JANZZon! can of course never comprehensively cover all terms, industries and languages, in specific cases we also deploy classic search mechanisms in the field of full-text search.
The structuring, the continuous expansion and maintenance as well as the updating and refinement of highly complex ontologies such as JANZZon! make it necessary to deploy expert tools, for example ontology editors, and require long-standing experience and a comprehensive set of skills within the field of knowledge modelling. Because such work, which tends to be complex and time consuming, can seldom be covered by in-house resources, JANZZon! with its cloud solution and its many individual customization options – including entirely private ontologies – offers the optimum and most cost-effective solution for your search and matching application.
Some of the key characteristics of JANZZon!
- Semantic, language-sensitive query completion: conventional search engines tend to offer only syntactical query completion; thanks to semantic contextual knowledge, JANZZon! is able to provide content-related query completions.
- Synonyms, different spelling (e.g. «-iza» versus «-isa»), abbreviations, hypernyms and hyponyms (parent-and-child concepts), and associated terms (e.g. «same, but different» or «part equivalent» concepts) are also taken into account
- High-performing full-text search as fallback in the event that JANZZon! is unable to provide relevant contextual knowledge
- Intelligent weighting of terms and concepts, for example using occupation classes (currently OCs 1 to 7); structural information is used in order to allocate different weightings to the occurrence of search terms in different areas of the dataset and in different contexts, for example in connection with the required occupation
- REST-based web-service interfaces
If required, JANZZon! can be customized flexibly and in accordance with the individual requirements of specific applications as well as in conjunction with JANZZsme! For a broader overview of the key benefits of semantic searches using JANZZon!, ask for our white paper «Keyword vs. ontology based, semantic Matching» (available in English only).