Voice recognition and digital technology
Chatbots capable of handling both text and voice are now everywhere. Support centers can provide service 24/7 thanks to these virtual assistants. Facebook provides new services with more than 11,000 chatbots in Messenger. Alexa at Amazon, Siri at Apple, Cortana at Microsoft and Google Home are now part of users’ daily lives.
Chatbots are capable of understanding voice and spoken language, which is the oldest form of communication. And thanks to Machine Learning, advances in voice recognition have been spectacular. In August 2017, Microsoft announced a rate of error of only 5.1% in voice recognition. This is the same rate of error as voice recognition by humans! Google is providing speech to text transcription in over 80 languages with Google Cloud Speech. New research is now focusing on noise reduction, the ability to understand several users at once, simultaneous translation and contextualization.
Semantics between human language and computer language
The ability to understand natural language requires the use of semantic analysis. NLP (Natural Language Processing) is used by computers and machines to decipher text entered directly or transcribed from speech. NLP corrects the text, analyzes sentence structure, recognizes the concepts expressed, analyzes the relationships between the concepts and deduces intent and components such as dates, names and places.
It also analyzes the humor, feelings and intent behind the words. These operations performed by NLP are based on a new type of knowledge, K-Data* (or Knowledge Data) which covers all the data processed by cognitive systems. This knowledge includes glossaries, synonyms, thesauruses, ontologies, scenarios and models of phrases. Through semantic analysis, K-Data reveals the meaning of the data within a specific field which affects the vocabulary recognized and determines the scope of possible actions.
Language analysis and actions in HR
Piles of documents to read, transcribing reports or interviews, voice commands in the field… The possible uses of natural language are endless. Concrete examples are beginning to emerge in HR and chatbots are becoming a reality. They help users to navigate or find information in HR computer systems. They guide applicants in completing their applications. They offer services to meet requests expressed in natural language such as requesting leave, updating details or requesting a certificate.
Semantic analysis allows you to imagine new applications and use HR information sources that have up to now not been considered due to their volume. You can now easily add information taken from job descriptions or applicants’ CVs to a skills database. Semantic understanding of titles allows you to group similar skills together or add emerging skills to the company’s database.
Semantic analysis is also useful when analyzing a company’s social climate. It helps analyze employees’ moods as well as their interests and concerns. Informal exchanges between employees, comments in surveys or interviews can all provide information that describes a company’s perception of wellbeing at work and the quality of its management. HR departments have access to new key indicators for determining social action and employee engagement.
Semantic analysis lets you make the most of new sources of information and decipher the meaning of data. HR can thus avoid the restrictive tasks and procedures related to the HR information system and feel free to focus on more important tasks. Semantic analysis provides the meaning inherent in words and is thus a major asset for HR.
* K-Data or Knowledge Data is a Sopra HR concept