Patent Application
20230014572
This application relates generally to human resource management. The application relates more particularly to an artificial intelligence, machine learning system that predicts happiness levels of employees by analyzing their attire relative to patterns and properties to provide recommendations to promote job retention and satisfaction.
Employee turnover is costly. Not only can it affect company financials, but it can also impact morale of remaining employees. This may affect productivity levels of those employees. Companies may invest in numerous programs to keep their employees happy in an attempt to reduce turnovers. Typical programs are generic and applied across the company, with little or no employee personalization. Routine inspections through traditional methods, like face to face meetings with management leaders, can provide insights to the employee's happiness. However, it is done infrequently or seasonally and prone to bias. Such meetings may have an opposite effect, causing stress to employees.
Various embodiments will become better understood with regard to the following description, appended claims and accompanying drawings wherein:
The systems and methods disclosed herein are described in detail by way of examples and with reference to the figures. It will be appreciated that modifications to disclosed and described examples, arrangements, configurations, components, elements, apparatuses, devices methods, systems, etc. can suitably be made and may be desired for a specific application. In this disclosure, any identification of specific techniques, arrangements, etc. are either related to a specific example presented or are merely a general description of such a technique, arrangement, etc. Identifications of specific details or examples are not intended to be, and should not be, construed as mandatory or limiting unless specifically designated as such.
Example embodiments herein disclose an artificial intelligence (artificial intelligence) system that tracks an employee's happiness by analyzing outfits that are worn by the employee to provide a happiness score for the employee, suitably on a daily basis. Should a happiness score fall below a predetermined threshold, the artificial intelligence system notifies management leaders and provides a recommendation of corrective measures to improve the happiness score through personalized rewards. Personalized rewards are selected to boost employee happiness.
The artificial intelligence system includes an image recognition component, happiness scoring component and recommendation engine component. The image recognition component utilizes computer vision to process the images from an office building security system. The security system captures imagery of employees entering and leaving the office building, and sometimes within the office building floor space as well. Each employee's access badge allows the cataloging of the entry by date and time. Image recognition detects clothing objects worn by the employee and the system classifies the objects with the corresponding employee. Object classification is not limited to labeling of the objects but also provides metadata for the objects such as coloration, newness, size, style, branding, and fit-finish detection. The additional metadata provides information to construct an outfit ensemble. Image recognition suitably uses crowdsourcing data to further build datasets from images scraped from popular fashion websites and events that would generally have happy people using the same or similar looks. Additional external datasets are suitably used for comparing and assisting in classification and continuous learning by the machine learning models in the happiness scoring component.
The happiness scoring component is suitably comprised in a series of machine learning features. Such features are algorithms that provide pattern recognition from the image recognition component datasets and outputs a percentage scoring factor if the employee is happy, suitably on a scale, such as a scale of 1 to 100, with 100 being 100% completely happy. A score is suitably sent to an administrator to review, and possibly initiate remedial action.
The recommendation engine component is suitably triggered when a score is below a predetermined threshold level and an occurrence frequency is higher than the predetermined threshold level. The recognition engine suitably integrates with an employee recognition system to select items that correlates with the employee's fashion attire. For example, if an employee often wears golfing attire, selected items may be golf balls and tees that are personally identified as rewards for the employee.
The artificial intelligence system is suitably trained using historical dataset from the security system and Human Resources system. The Human Resources system provides historical dataset of the traditional face to face interviews, the length of time an employee stays on the job, attendance, tardiness, self-reporting surveys, and companywide surveys. The artificial intelligence system is suitably evaluated using specified targeted employees and then inspected using traditional face to face interviews to confirm happiness levels and recommendations provided.
The artificial intelligence system suitably uses supervised learning to analyze data points from employee clothing to predict the happiness level of the employees. The artificial intelligence system generates a personalized recommendation to cheer up an employee to improve and or maintain a happiness level that reduces turnover.
Crowd-sourced data, suitably obtained from freely available images of clothes shown in popular fashion websites, serves to continuously evaluate machine learning models to detecting fit and finish, newness, branding, and concepts such as “dress to impress.” Crowd-sourced data may also include data obtained from freely available images of clothes worn by people at happy events to continuously evaluate the machine learning models on detecting the outfit attire that depicts happiness
Referring now to the drawings,
In a particular example, when employee 106 enters business premises 104 at entryway 114, the employee is wirelessly identified by his company ID badge 128 via wireless receiver 136. An image of the identified employee is captured by camera 112. Employee information and captured image data is communicated to server 120 via network cloud 140. As will be detailed further below, captured image data is subjected to analysis by application of artificial intelligence (AI) and machine learning (ML) by server 120, suitably including additional information. Such additional information may include human resources information, such as employee badge information, personal information, attendance record, tardiness record, employment duration, birthday, and the like. Additional information suitably implemented includes results from self-reporting or companywide services, employee ID data, additional image captures or a company holiday calendar. Fashion image data is also suitably implemented for analysis and recommendations, including information from Internet sites such as fashion site images or crowdsourcing data.
Once an employee image has been analyzed, recommendations to improve employee happiness are generated. A recommendation may include a personal or personalized employee gift corresponding to analysis of a captured employee image. In the illustrated example, employee 106 is determined to be in golfing attire and a recommended gift includes golf balls and golf tees. A corresponding order is suitably placed, automatedly or manually by an administrator, at online store 144. Online store 144 can then ship ordered goods for delivery to John Doe, illustrated at 106′.
Image 204, supplemented by additional information 220, forms image dataset 232. Image dataset 232 includes images such as images 236, 240, 244 and 248. Dataset images suitably include extracted patterns, such as pattern 252 from image 236 and patterns 250 and 251 from image 240. Image patterns are suitably associated with other image patterns, additional information or employee recommendations, such as illustrated by associative ordering 253 associated with image 244. Image dataset 232, illustrated at 232′, is analyzed at block 256, along with human resources information 262, to form an evaluation dataset 260, determine an employee happiness index 264, and a suggestion 270.
Turning now to
Image recognition data is stored at block 540. Stored data, along with human resource pipeline data 544, provides happiness scoring at block 548. Such scoring directs interactive training at block 552, and generation and management of alerts at block 556.
Happiness scoring allows for a determination of an employee happiness level at block 560. If an employee is determined to be happy, a prediction iteration ends at block 564. If an employee is determined to be unhappy at block 560, such as having a happiness index below a preselected threshold level, a recommendation is generated at block 564, which may include engaging reward systems pipeline 568, before proceeding to alert generation and management at block 556.
While certain embodiments have been described, these embodiments have been presented by way of example only, and are not intended to limit the scope of the inventions. Indeed, the novel embodiments described herein may be embodied in a variety of other forms; furthermore, various omissions, substitutions and changes in the form of the embodiments described herein may be made without departing from the spirit of the inventions. The accompanying claims and their equivalents are intended to cover such forms or modifications as would fall within the spirit and scope of the inventions.
