Industry Research: Machine Learning in HR Can Be Transformative
Be sure your data is ready
Posted on 12-23-2019, Read Time: Min
Share:
New applications of machine learning in HR are about to change the profession in fundamental ways.
A European financial institution needed to comb through their human resources files to identify personal information in order to comply with new privacy regulations. Since the task was impractically large for humans, the company became interested in utilizing machine learning (ML), a type of artificial intelligence (AI) software.
As they continued to scope the possibility of utilizing ML, the company’s HR professionals realized they could gain ancillary benefits by creating predictive hiring models based on archived data. The ML team taught the algorithm to scour five years of employee data to correlate the backgrounds and skills that best predicted job candidates’ success on the job. The story is a good example of how the HR function is being transformed by AI technology.
Changes on the HR Horizon
HR pros see it coming — a survey conducted by HR.com of its members found that 86% believe their profession will be transformed over the next five to 10 years. According to HR.com’s “The 2019 State of Artificial Intelligence in Talent Acquisition” report, 64% of HR professionals expect to apply AI to recruitment.
The field is ripe for automation. Much of the work that HR organizations do, while important, is repetitive, such as processing and filing forms, scanning résumés and complying with information requests. Many of those tasks could be offloaded to computers, freeing humans for more analytical tasks.
Recruitment is a particularly choice area for automation. “Machine learning is targeted at learning for massive amounts of data and spotting patterns,” said Anke Conzelmann, director of product management at Iron Mountain in a recent webcast entitled The Advancing HR Function: The Future of HR and HR Technologies. “It can be used to find people based upon data rather than guesswork.”
The field is ripe for automation. Much of the work that HR organizations do, while important, is repetitive, such as processing and filing forms, scanning résumés and complying with information requests. Many of those tasks could be offloaded to computers, freeing humans for more analytical tasks.
Recruitment is a particularly choice area for automation. “Machine learning is targeted at learning for massive amounts of data and spotting patterns,” said Anke Conzelmann, director of product management at Iron Mountain in a recent webcast entitled The Advancing HR Function: The Future of HR and HR Technologies. “It can be used to find people based upon data rather than guesswork.”
Get to Work
The survey found that many HR organizations feel unprepared for the technology that will transform the function. While nearly half of the HR professionals surveyed give their departments high marks at meeting the current needs of the organization, only 36% said their group was prepared to thrive over the next three to five years.
Algorithms are more effective than people at correlating factors from a candidate’s background with successful employees’ profiles and matching people to the right jobs, Conzelmann said. But HR professionals should avoid the temptation to see machine learning in HR as a black box.
Algorithms are more effective than people at correlating factors from a candidate’s background with successful employees’ profiles and matching people to the right jobs, Conzelmann said. But HR professionals should avoid the temptation to see machine learning in HR as a black box.
Importance of a Strong Data Foundation
You need a strong data foundation to successfully implement new technologies successfully. “It’s very important to understand that machine learning learns from the data you feed it,” Conzelmann said. “If you give it bad data, it’s going to find patterns based upon that bad data.” That makes data preparation a critical step.
Before applying automation, the HR organization first needs to determine the objective. “Write down a list of the questions you would love to answer, then walk backwards and determine what data and documents contain that information,” she said. To assess candidates, for example, the organization needs to gather all historical data that’s relevant to the recruiting process, such as customer relationship management profiles, internal reviews, and education and certification records. The more data the machine has, the better.
Many historical records may be on paper, so a third-party service can be useful in digitizing those documents and extracting relevant information. Data should be classified and tagged, so the machine knows what to look for and can help in future retrieval.
Beware of introducing unintentional bias, Conzelmann advised. For example, if an organization has had difficulty retaining talent in the past, relying strictly upon historical records to train the algorithm could backfire. “If I’ve made bad decisions in the past, training the machine to make the same bad decisions isn’t going to move me forward,” she said. “Be sure you have the breadth of data so the machine can look for the right patterns.”
Conzelmann and Iron Mountain Senior HR Director Ellen Donovan shared these tips for HR transformation with AI:
Before applying automation, the HR organization first needs to determine the objective. “Write down a list of the questions you would love to answer, then walk backwards and determine what data and documents contain that information,” she said. To assess candidates, for example, the organization needs to gather all historical data that’s relevant to the recruiting process, such as customer relationship management profiles, internal reviews, and education and certification records. The more data the machine has, the better.
Many historical records may be on paper, so a third-party service can be useful in digitizing those documents and extracting relevant information. Data should be classified and tagged, so the machine knows what to look for and can help in future retrieval.
Beware of introducing unintentional bias, Conzelmann advised. For example, if an organization has had difficulty retaining talent in the past, relying strictly upon historical records to train the algorithm could backfire. “If I’ve made bad decisions in the past, training the machine to make the same bad decisions isn’t going to move me forward,” she said. “Be sure you have the breadth of data so the machine can look for the right patterns.”
Conzelmann and Iron Mountain Senior HR Director Ellen Donovan shared these tips for HR transformation with AI:
- Start exploring the potential of using data more strategically in HR.
- Understand what data you have and build the skills to manage it effectively.
- Apply retention policies, particularly in regulated environments. “Keeping everything forever is no longer an option,” Conzelmann said.
- Enrich once, use repeatedly. Once data has been cleaned and prepped, find as many valuable uses as possible across the organization.
- Think of AI as an assistant, not a replacement. “While AI can provide a lot of insights quickly, it does not have that common sense that a person has,” Donovan said. “It has to be human plus AI.”
Author Bio
 |
Paul Gillin is a writer, speaker and B2B content marketing strategist who specializes in social media and IT. Previously, Gillin was a technology journalist for 23 years. He was founding editor-in-chief of B2B technology publisher TechTarget and editor-in-chief and executive editor of the technology weekly Computerworld. Gillin is the author of five books and more than 400 articles on the topic of social media and digital marketing. He is currently a staff columnist at Biznology.com. Visit www.gillin.com Connect Paul Gillin |
Error: No such template "/CustomCode/topleader/category"!
