The dynamics of the labor market and the tasks with which jobs are being composed are continuously evolving. Job mobility is not evident, and providing effective recommendations in this context has also been found to be particularly challenging. In this paper, we present Labor Market Explorer, an interactive dashboard that enables job seekers to explore the labor market in a personalized way based on their skills and competences. Through a user-centered design process involving job seekers and job mediators, we developed this dashboard to enable job seekers to explore job recommendations and their required competencies, as well as how these competencies map to their profile. Evaluation results indicate the dashboard empowers job seekers to explore, understand, and find relevant vacancies, mostly independent of their background and age.
Exploring Job Recommendations with a Visual Dashboard
1. Explaining and Exploring Job
Recommendations: a User-driven Approach
for Interacting with Knowledge-based Job
Recommender Systems
Francisco Gutiérrez, Sven Charleer, Robin De Croon, Nyi Nyi Htun, Gerd
Goetschalckx, Katrien Verbert
francisco.gutierrez@cs.kuleuven.be
@FranciscoGhz
1
katrien.verbert@cs.kuleuven.be
@katrien_v
AUGMENT
http://augment.cs.kuleuven.be/
2. 22
• Abundant overload of job vacancies
• Dynamic Labor Market: need to support job mobility
• Providing effective recommendations particularly
challenging.
• Need for:
increased diversity
explanations
user control
exploration
Problem: interaction with job RecSys
needed
4. Research Questions
4
[RQ1] Does enabling job seekers to interact with
visualization techniques empower them to explore,
understand, and find job recommendations?
[RQ2] Do personal characteristics, such as age and
background, impact the user perception and user
interaction with such an interface?
5. 5
Interactive & Job Recommender Systems
SetFusion (Parra et al., 2014)
LinkedVis (Bostandjiev, 2013)
JobStreet (Bakri et al., 2017)
11. Labor Market Explorer Design Goals
11
[DG1] Exploration/Control
Job seekers should be able to control
recommendations and filter out the
information flow coming from the
recommender engine by prioritizing specific
items of interest.
[DG2] Explanations
Recommendations and matching scores should
be explained, and details should be provided
on-demand.
[DG3] Actionable Insights
The interface should provide actionable
insights to help job-seekers find new or more
job recommendations from different
perspectives.
24. Discussion
[RQ1] User Empowerment
• The approach is perceived as effective to explore job
recommendations.
• Most participants felt confident and will use the explorer again.
• Explanations contribute to support user empowerment.
• A diverse set of actionable insights were also mentioned by
participants.
24
25. [RQ2] Personal Characteristics
• The explorer was slightly better perceived by older participants (45+).
• Participants in the technical group engaged more with all the different
features of the dashboard.
• Non-native speakers, sales and construction groups engaged more with
the map.
• The table overview was perceived as very useful by all user groups, but
the interaction may need further simplification for some users.
Discussion
25
26. • A visual representation of the labor market was deemed
interesting by both job seekers and job mediators.
• Tool inspired by the UpSet visualization technique.
• Competence-based explanations, the diverse set of filters,
location-based search and overview features, were
relevant.
Design Implications
26
27. User-centered design process involving both job seekers
and job mediators
Key features:
Design Implications
27
• Overview first, favorite competences of interest.
• Competence-based explanations.
• Actionable, job market related insights.
• Diverse set of filters.
28. • Future work will focus on a “simulation mode”.
• Further investigate job mobility scenarios.
• Explore “What-If” scenarios.
• Autonomous exploration of the job market.
Future Work
28
29. Explaining and Exploring Job
Recommendations: a User-driven Approach
for Interacting with Knowledge-based Job
Recommender Systems
Francisco Gutiérrez, Sven Charleer, Robin De Croon, Nyi Nyi Htun, Gerd
Goetschalckx, Katrien Verbert
29
AUGMENT
francisco.gutierrez@cs.kuleuven.be
@FranciscoGhz
katrien.verbert@cs.kuleuven.be
@katrien_v
http://augment.cs.kuleuven.be/
Editor's Notes
Good afternoon, it is our/my pleasure to present our work on explaining and exploring job recommendations that we elaborated in collaboration with VDAB, the public employment service of Flanders.
[TODO] add VDAB logo and ESF logo
Job recommeder systems are increasingly researched: as in many application domains, there is an abundant overload ofjob vacancies that need to filtered out.
In additon, the application domain faces some other challenges: the dynamics of the job market and the tasks with wich jobs are composed are rapidly evolving and job seekers are expected to be able to deal with these changes in an efficient manner, and to move easily within this transitional job market.
Such Job mobility is not evident: many jobs actually require a similar skill set, but job seekers are often not aware of the overlap between required competences in different job types.
As a result, providing effective recommendations has been found to be particularly challenging. There is a need to increase the diversity of recommended jobs, to explain the recommendations so that job seekers can better understand why a job they would not expect to have a good competence match for is recommended and to support better user control and exploration
We tackle these challenges by combining recommendation with visualisation techniques to explain job recommendations: visualisation techniques explain which tasks or competences are required for a vacancy and give actionable insights
In addition, we research how we can use interactive visualistation techniques to support exploration and user control over a broad set of job recommendations, so that job seekers can explore a very broad and diverse set of recommendations and filter out the more relevant ones with interaction techniques
The research questions that we address are: [READ RQs]
Our work is inspired by tools liked LinkedVis that visuale the links between skills, degrees, connections and recommendations and that also enable end-users to tweak the importance of input parameters like the skills that are required.
But also by tools like setfusion that visualise the relationship between input parameters and recommendations: you can for instance see which recommendations match multiple input parameters by looking at the intersection
We also looked at more geberal job applications that visualise available vacancies on a map.
Using a user centered design approach, we then elaborated an interactive dashboard on top of the knowledge-based recommender system of VDAB.
The objectives are to explore and explain a broad and diverse set of job recommendations
And to support actionable insights to job seekers to potentially support job mobility
This is an overview of the user-centered design methodology that we used. We conducted two focus groups with job seekers and job mediators. We started from the parameters that the knowledge-based ELISE recommender system uses and asked participants to rank the importance of these parameters.
We then elaborated initial designs and conduced to co-design sessions, again with both job seekers and job mediators.
Then there were two iterations of evaluations with first prototypes where we used the think aloud protocol to capture feedback.
And finally we did a final evaluation with 66 job seekers with the final prototype.
These are the parameters that are used in the knowledge-based recommender system of VDAB: examples are job title, completencies, work experience. In the focus groups, participants ranked the parameters.
The results are presented here: job title and distance to a job were ranked frequently on the first positiion by both job seekers and job mediators. Many other parameters were ranked frequently on the next positions, including type of contract, competences, studies and work experience.
We then elaborated designs that put focus on the important parameters: location and distance are for instace visualised in all tthese design, as well as competences that are asked in jobs. We captured additional input of job seekers and mediators in co-design sessions and evaluated first prototypes.
Results of these sessions were captured and articulated in the following design goals:
Exploration / control: the need to control recommendations and highlight items of interest was expressed
Job seekers also wanted to see explanations of recommendations that go beyong the typical matching score
There is also a need to support actionable insights: for instance by visualising which competences are currently in high demand in the job market
This is the final design of the dashboard we elaborated. This is the overview visualisation that visualises job recommendations, their required competences and a map overview that provides distance or location insights
The overview visualisation can then be expanded to see the details of the compentences and the potential competence gap
To support Design goal 1, there are various filters that enable job seekers to narrow down the initial broad range of vacacies, for instance by type of contract or location.
Recommendations are explained by visualising the required competences: the view on top is an overview visualisation that can then be expanded.
a blue dot indicates that the job seeker masters the compentence and an empty dot indicates that the competence is requested but not yet mastered.
In the detailed view, job seekers can also filter out skills they are not interested in or they can move skills of interest to the front of the table by starring them.
Inspired by the Upset Visualisatiion techniques, we show with miniaturized bar charts on top that indicates how often a particular skill or competence is asked for in job recommendations, and this can provide additional actionable insights into the job market
We conducted a final evaluation with 66 job seekers, in different training programs at VDAB – including highly technical users, sales, construction and non-native speakers.
We asked the users to freely explore the dashboard and then used the ResQue questionnaire to capture user experience feedback together with two open questions.
We also logged all interactions.
These are the results of the ResQue questionnaire: overall the dashboard was well received – particularly also the explanations and ease of use.
This positive trend was maintained across the different backgrounds. Technical users were a bit more negative with respect to overall satisfaction (Q5). They gave very detailed and concrete suggestions of how particular components could be further improved, including resizability of the table.
Perceived accuracy (Q1) was a bit lower for non-native speakers: As language is not trivial for this group, also working with an interface in a different language as well as finding suitable jobs is evidently much more difficult.
Perceived usefulness (Q2) was a bit lower for the construction group. The wide range of very specific competencies in their domain was one of the reasons they highlighted for this lower score.
In general, we do not see many differences across the different age categories. Overall, all questions were answered positively in the different age group. The explorer was slightly better perceived by older participants (45+) with respect to overall satisfaction (Q5) and use intention (Q4). Confidence (Q6) was slightly higher for younger participants.
To better understand the use of the tool, we logged the clicks of participants through the different visualization components. – including the vacancies map, the filters and the vacancies table
Participants with technical background engaged the most with the job vacancies table, where they performed most of the interactions in a significant different way compared to the other groups.
Participants in the construction group and non-native speakers engaged less with the job vacancies table. We can observe that these groups engaged more with the map component.
Technical users frequently used the multiple features of the table, including the favorites, sorting, and filtering features, with a significant difference compared to the other. Although competence-based sorting was not used frequently by the other groups, we do see that the sales group also engaged frequently with the favorites option.
Responses to the questions indicate that job seekers value the use of interactive visualization techniques to find relevant job vacancies.. In general, the approach is perceived as effective to explore job recommendations. The interaction patterns also indicate that participants engaged well with the interface.
Also explanations seem to contribute well to better support of user empowerment. Many of the responses hint to better understanding of the job recommendations by being able to see which competencies are required.
A diverse set of actionable insights were also mentioned by participants. Participants indicate that the overview of competencies enables them to explore the job market and understand whether they have the needed competencies. They also gain insight into regions where most jobs are offered, supporting potential location-based job mobility. Other actionable insights include the observation that their user profile is not up to date. Such insights are key as well, as they trigger job seekers to update their profile and, in turn, receive better recommendations.
In general, we can observe only few differences with respect to the age and gender of different participants. The explorer was slightly better perceived by older participants who all indicated that the explorer is a good tool and that they would use the explorer again.
We can highlight the importance of a user-centered design process involving both job seekers and job mediators
The iterative design process identified several key features for interacting with an overview of a broad range of job recommendations, including the option to favorite competencies that are of interest. The difficulty here is that there is often a very large list of competences that needs to narrowed down to a manageable overview.
In addition, a visual representation of labor market related information such as the number of times a specific competence is demanded was deemed interesting by both job seekers and job mediators, and was conceptualized as a miniature bar chart in the dashboard, inspired by the UpSet visualization technique
Good afternoon, it is our/my pleasure to present our work on explaining and exploring job recommendations that we elaborated in collaboration with VDAB, the public employment service of Flanders.
[TODO] add VDAB logo and ESF logo