The COVID-19 pandemic has affected the lives and health of many people. Among them, health care workers indeed played the biggest role. In this work, we investigate the health care workers job applications, and the job postings change with the time series analysis perspective. The dataset is provided by Kariyer.net, which is the foremost job-searching company in Turkey. Since the data set includes job applications over every region of country, the results reveal important facts about health care workers in Turkey. We aim to study the possible effect of vaccination on health care workers job search decisions.

We examine the principles of Benford’s Law on a data set of the daily number of COVID-19 cases and deaths per country. Benford’s Law was used before on the coronavirus-related data sets to find anomalies; however, our work differs from existing ones by first using the latest data set covering almost 3 years of coronavirus cases and second, proposing a new analysis, checking monotony, for Benford’s Law. First, the chi-square test was performed. The expected percentages for each case were calculated according to Benford’s Law and compared to the observed values. Second, the mean absolute distance score was also calculated for each country. These two analyses gave similar results: many countries in various regions of the world were well above the threshold value. Third, it was examined whether the frequencies of the first digit were monotonically decreasing. The analysis of all the results obtained reveals that Benford’s Law is applicable to health-related data analysis, as it was previously used in financial or network data sets.

The term "sector" in professional business life is a vague concept since companies tend to identify themselves as operating in multiple sectors simultaneously. This ambiguity poses problems in recommending jobs to job seekers or finding suitable candidates for open positions. The latter holds significant importance when available candidates in a specific sector are also scarce; hence, finding candidates from similar sectors becomes crucial. This work focuses on discovering possible sector similarities through relational analysis. We employ several algorithms from the frequent pattern mining and collaborative filtering domains, namely negFIN, Alternating Least Squares, Bilateral Variational Autoencoder, and Collaborative Filtering based on Pearson’s Correlation, Kendall and Spearman’s Rank Correlation coefficients. The algorithms are compared on a real-world dataset supplied by a major recruitment company, Kariyer.net, from Turkey. The insights and methods gained through this work are expected to increase the efficiency and accuracy of various methods, such as recommending jobs or finding suitable candidates for open positions.

Most autoimmune diseases are generally influenced by HLA genes, which play a role in their onset and progression. In most cases, they are not counted as a diagnostic criterion for autoimmune diseases, but they act as a predisposing agent for the onset of the disease. The goal of this research is to look at the link between disease, genetics, and ethnicity. By clustering the HLA genes predisposing to selected autoimmune diseases using different clustering algorithms, we isolated the clusters that contained the samples having predisposing genes. After carrying out a homogeneity analysis, we arrived at the results concerning the ethnicity of the samples predisposed to the corresponding diseases. A well-known validation metric, the silhouette coefficient, is used to assess the methods. The accuracy of the findings is determined by comparing final clusters of HLA genes and the literature on the epidemiology of selected autoimmune diseases. The results demonstrate that clustering assists to find predisposing genes in distinct groups.

Overlapping community detection allows placing one node to multiple communities. Up to now, many algorithms are proposed for this issue. However, their accuracy depends on the overlapping level of the structure. In this work, we aim at finding relatively small overlapping communities independently than their overlapping level. We define k-connected node groups as cohesive groups in which each pair of nodes has at least k different node disjoint paths from one to another. We propose the algorithm EMOC first finding k-connected groups from the perspective of each node and second merging them to detect overlapping communities. We evaluate the accuracy of EMOC on artificial networks by comparing its results with foremost algorithms. The results indicate that EMOC can find small overlapping communities at any overlapping level. Results on real-world network show that EMOC finds relatively small but consistent communities.

Many methods have been proposed to detect communities in complex networks, but very little work has been done regarding their interpretation. In this work, we propose an efficient method to tackle this problem. We first define a sequence-based representation of networks, combining temporal information, topological measures and nodal attributes. We then describe how to identify the most emerging sequential patterns of this dataset and use them to characterize the communities. We also show how to highlight outliers. Finally, as an illustration, we apply our method to a network of scientific collaborations.

Community detection consists in searching cohesive subgroups in complex networks. It has recently become one of the domain pivotal questions for scientists in many different fields where networks are used as modeling tools. Algorithms performing community detection are usually tested on real, but also on artificial networks, the former being costly and difficult to obtain. In this context, being able to generate networks with realistic properties is crucial for the reliability of the tests. Recently, Lancichinetti et al. designed a method to produce realistic networks, with a community structure and power law distributed degrees and community sizes. However, other realistic properties such as degree correlation and transitivity are missing. In this work, we propose a modification of their approach, based on the preferential attachment model, in order to remedy this limitation. We analyze the properties of the generated networks and compare them to the original approach. We then apply different community detection algorithms and observe significant changes in their performances when compared to results on networks generated with the original approach.