TiiQu suitability match is a two-sided algorithm that integrates questions from both the perspectives of an employee looking for an employer, and an employer looking for an employee. Optimally, both parties will have a strong interest in the mission to be completed. In other words, TiiQu-Match will be applicable to an employer seeking an expert for the completion of a specific mission either as an independent worker or as a team member within their own employees and vice versa. This way, TiiQu-Match will enforce employee-employer collaborations by minimizing uncertainties and facilitating smart contracts.
- Staffing challenges and online job recommendations
- Ethical Considerations
- Open Questions
Preliminaries and Definitions
There is a great deal of research in the staffing and recruiting area which provide insights into how various expert attributes relate to job success and performance. For instance, studies have shown that on a scale of 0-1, work sample correlates with performance with a score of 0.54. Interestingly, integrity/honesty (qualities which we separately quantify using the Trust Quotient) were found to be better predictors of performance (score of 0.41) than biographical information (education, training, work experience, interests), which scored only 0.35 . However, these findings have not been widely embraced by HR practitioners, possibly because of the extensive amount of time and effort required to isolate key job performance indicators from research studies. Consequently, employers often rely on their “gut instincts” and misconceptions in making hiring decisions, and this can be costly for organizations. For instance a study conducted with a sample of 5000 HR management members found that 72% agreed that conscientiousness is a better predictor of employee performance than intelligence, whereas research evidence suggest that this judgment may actually incorrect . Also in a recent survey of working U.S. adults in their 30s, 73% of respondents expressed some level of dissatisfaction with their jobs, reflecting a serious mismatch between career ambitions of employees and the jobs responsibilities they currently occupy . It is no secret that loss of interest and dissatisfaction are often associated with low performance on the job.
Staffing challenges and online job recommendations
A survey of 33,000 employers from 23 countries found that 40% of them had difficulties in finding and hiring the desired talent, and approximately 90% of nearly 7,000 managers indicated that talent acquisition and retention were becoming more difficult . As 21st century trends drive towards digitization, many online professional platforms have emerged with the goal to help professionals find employers and vice versa. Online professional platforms typical adopt a strategy based on Recommender Systems, in which some algorithm gathers information from a user’s profile and activities, and uses this information to recommend jobs that might be of interest to the user. Similarly, the engine can recommend user profiles which might be of interest to an employer. These systems are similar to the algorithm that Amazon uses to recommend items to their customer, or the Netflix movie recommender engine. For example, LinkedIn has a “Similar Profiles”  recommender which recommends candidates with similar profiles to employers after the latter have found a candidate they like. This is in the spirit of "users who watched that movie also watched these movies". While this approach is useful in some way, a number of very important questions are overlooked.
- First, every recommendation is made based on unverified information provided by the user, hence the question of whether the recommended candidate is trustworthy and actually fit for the job is never addressed.
- Second, recommending jobs to people is different from recommending movies. While it's OK to recommend movies to a user, it does not make sense to recommend users to movies. But such a bilateral approach is what is needed for job recommendations .
- Third, employers are always looking for the best match, so are employees. Hence, there must be a mechanism to perform fair and objective comparison among users before final recommendations are made.
The TiiQu Suitability Algorithm (TiiQu-Match)
The TiiQu Suitability algorithm is bilateral recommendation engine that aims to:
- Match candidates to tasks for which they possess the skills and abilities to complete.
- Match employees to employers (or teams) with whom they are able to collaborate successfully.
In this regard, the algorithm consists of a top-level recommendation engine (profile filter) to perform personalized recommendations based on skills, and the TiiQu-Match to perform a more screened pairing of employers with employees. Rather than simply filtering users profiles to make recommendations, TiiQu-Match searches for conditions that result in optimum chemistry between the employee and the employer. Hence, a relevant question for an employer could be phrased as follows: given x number of possible suitable candidates with different sets of skills, which choice of candidate(s) will minimize resources necessary to complete a defined task? Variables to minimize may include turn-over rate (for permanent position jobs), time of completion (for time-sensitive tasks), cost (for hourly jobs)? In fact, this is the very essence of interviewing. A similar question can be phrased for the employee. TiiQu-Match is meant to address these questions by considering trustworthiness (measured by the Trust Quotient), suitability (skills relevant for a given task), and chemistry (how well would the parties work together). It is also worth noting that TiiQu-Match for the moment focuses on job-matching within the STEM fields. Attempting to make recommendations for all possible areas of expertise and for users of all backgrounds significantly reduces the quality of the matching between experts and employers. The focus on STEM fields allows TiiQu-Match to avoid such a compromise.
There are ethical issues that must be taken into consideration when developing a job recommender engine. For instance, will it be ethically right if a feature like age is considered as a matching criteria for a given job? On one hand, It may be possible that certain age groups are more suitable for certain jobs, in which case the age factor will bring additional value to the recommendation engine. On the other hand, age discrimination is ethically wrong and must be avoided. This section will be expanded after discussions with ethical professionals.
- Can a personality test algorithm improve employee-employer matching?
- Have certain combinations of skills proven to be more effective certain jobs?
 : Ryan, Ann Marie, and Nancy T. Tippins. "Attracting and selecting: What psychological research tells us." Human
Resource Management 43, no. 4 (2004): 305-318.
 : Rynes, Sara L., Kenneth G. Brown, and Amy E. Colbert. "Seven common misconceptions about human resource practices: Research findings versus practitioner beliefs." The Academy of Management Executive 16, no. 3 (2002):92-103.
 : Stephens, Bryan, Rajiv Garg, and John Butler. "'Which Way I Ought to Go?” Career Trajectories of Entrepreneurs, Executives, and Senior Managers in High-Tech." (2018).
 : Ployhart, Robert E. "Staffing in the 21st century: New challenges and strategic opportunities." Journal of management 32, no. 6 (2006): 868-897.
 : https://www.quora.com/How-does-LinkedIns-recommendation-system-work
 : Jochen Malinowski and Tobias Keim: Matching People and Jobs: A Bilateral Recommendation Approach. Proceedings of the 39th Hawaii International Conference on System Sciences, 2006