Tamer N. Jarada

Link prediction and classification in social networks and its application in healthcare and systems biology

Prediction is one of the most attractive aspects in data mining. Link prediction has recently attracted the attention of many researchers as an effective technique to be used in graph based models in general and in particular for social network analysis due to the recent popularity of the field. Link prediction helps to understand associations between nodes in social communities. Existing link prediction-related approaches described in the literature are limited to predict links that are anticipated to exist in the future. To the best of our knowledge, none of the previous works in this area has explored the prediction of links that could disappear in the future. We argue that the latter set of links are important to know about; they are at least equally important as and do complement the positive link prediction process in order to plan better for the future. In this paper, we propose a link prediction model which is capable of predicting both links that might exist and links that may disappear in the future. The model has been successfully applied in two different though very related domains, namely health care and gene expression networks. The former application concentrates on physicians and their interactions while the second application covers genes and their interactions. We have tested our model using different classifiers and the reported results are encouraging. Finally, we compare our approach with the internal links approach and we reached the conclusion that our approach performs very well in both bipartite and non-bipartite graphs.

Applications of Social Network Construction and Analysis in the Medical Referral Process

The application of social network analysis (SNA) and mining in health care domains has recently received a considerable attention for its key role in understanding how doctors form communities, and how they are socially connected with each other. This understanding helps enhance organizational structures and process flows. In this paper, we show how SNA techniques can solve issues in the medical referral system by analyzing the social network of general practitioners (GPs) and specialists (SPs) associated with a medical referral system in the Canadian healthcare system and the like. Various SNA and mining procedures are proposed backed by experimental results.

MCF: A tool to find multi-scale community profiles in biological networks

Recent developments of complex graph clustering methods have implicated the practical applications with biological networks in different settings. Multi-scale Community Finder (MCF) is a tool to profile network communities (i.e., clusters of nodes) with the control of community sizes. The controlling parameter is referred to as the scale of the network community profile. MCF is able to find communities in all major types of networks including directed, signed, bipartite, and multi-slice networks. The fast computation promotes the practicability of the tool for large-scaled analysis (e.g., protein-protein interaction and gene co-expression networks). MCF is distributed as an open-source C++ package for academic use with both command line and user interface options, and can be downloaded at http://bsdxd.cpsc.ucalgary.ca/MCF. Detailed user manual and sample data sets are also available at the project website.

Mapping rules for converting from ODL to XML schemas

This paper presents a comprehensive approach for the transformation of ODL Schemas into XML Schemas. The approach starts with an incomplete set of rules described in the literature to assist in the transformation process. The fact that the rules provided a solid foundation for expansion, as well as the fact that the rules only cover a small subset of ODL, was our main motivation for continuing the study of this topic. In this paper, we first analyze an existing set of nine transformation rules. After evaluating the correctness and completeness of the rules, we proceed to propose some improvements and extensions into a more complete set of rules that cover the whole transformation process. By modifying the existing rule set, we are able to handle a much wider variety of ODL. Finally, we discuss some ODL scenarios that the original rule set cannot handle. This is meant to justify the need for the proposed extension as described in this paper. The presented more complete rule set is capable of handling a larger subset of ODL (including dictionaries, global and local scope enumerations, and most importantly, inheritance).

Robust framework for recommending restructuring of websites by analysing web usage and web structure data

The work described in this paper is motivated by the fact that the structure of a website may not satisfy a larger population of the visiting users who may jump between pages of the website before they land on the target page(s); this is at least partially true because access patterns were not known when the website was designed. We developed a robust framework that tackles this problem by considering both web log data and web structure data to suggest a more compact structure that could satisfy a larger user group. The study assumes the trend recorded so far in the web log reflects well the anticipated behaviour of the users in the future. We separately analyse web log and web structure data using three techniques, namely clustering, frequent pattern mining and network analysis. The final outcome from the two stages is reflected on to one of the six models, namely the network of pages to report linking pages by the most appropriate connections.

A novel client-based approach for signing and checking web forms by using XML against DoS attacks

In parallel to rapid growth of internet technologies, security becomes more critical in various real life applications such as e-finance, e-health, and e-government. These applications strictly require data authentication mechanisms. To address this essential issue, we grasp the idea of client based authenticity for interactive web technologies. We proposed a novel client based web form signing and checking with XML data structure method. Our method specifically uses XML structure for the involvement of data exchange between web applications. Our method curbs the DoS (Denial of Service) attacks for protection of the server. In order to illustrate our ideas, we adapted our digital signature mechanism on health related forms with two commonly used web browsers.

Link prediction and classification in social networks and its application in healthcare

Prediction is one of the most attractive aspects in data mining. Link prediction has recently attracted the attention of many researchers as an effective technique to be used in social network analysis to understand the associations between nodes in social communities. It has been shown in the literature that the link prediction technique is limited to predict the existence of the links in the future. To the best of our knowledge, none of the previous works in this area has explored the prediction of the links that could disappear in the future. In this paper, we propose a link prediction model that is capable of predicting link that might exist and links that may disappear in the future. The model has been successfully applied in two different domains, namely health care and stock market. We have tested our model using different classifiers and the reported results are encouraging.

Time Series Analysis for the Most Frequently Mentioned Biomarkers in Breast Cancer Articles

Breast cancer biomarkers have received a considerable attention for their key role in detecting and preventing the causes of breast cancer. In this paper, we study the impact of the published research related to the top genes most frequently mentioned in breast cancer articles. Our study helps governments and organizations by giving an idea about the number of studies that probably needs to be targeted in their support and funds. We perform time series analysis for the most frequently mentioned biomarkers in breast cancer articles. Constructing our time series dataset involves Information Retrieval (IR), Entity Recognition (ER) and Information Extraction (IE). We build a time series for the most frequently mentioned biomarkers in breast cancer articles by computing the number of published articles that mentioned these biomarkers over a periodic period of time. We use the autoregressive moving average (ARIMA) to build a model that helps in understanding and predicting a future number of articles in the time series of the breast cancer biomarkers.

Integrating Online Social Network Analysis in Personalized Web Search

With the emergence of high speed internet applications and advanced Web 2.0 based Rich Internet Applications (i.e., blogs, wikis, etc.), it has become much easier for the users to publish data over the Web. This brings a challenge for the Web search solutions to let individual users find the right information as per their preferences, because traditional Web search engines have been built on “one size fits for all” concept. Different users of the Web may have different preferences. Search results for the same query raised by different users may differ in priority for individual users. In this book chapter, we present the extended version and results of our proposal on community-aware personalized Web search. It is quite challenging to know the preferences of the users by the search engines. We have designed and developed our unique approach of finding the preferences of users from the relevant parts of the user’s social network and community. We believe that the information related to the queries posed by the users may have strong correlation with the relevant information in their social networks. In order to find out personal interest and social-context, we find (1) activities of users in their social-network, and (2) relevant information from user’s social networks, based on our proposed trust and relevance matrices. We have further developed a mechanism that extracts from user’s social network information to be used to re-rank search results from a search engine. We also have discussed the implementation and evaluation details of our proposed solution.

Data Analysis Based Construction and Evolution of Terrorist and Criminal Networks

The wide-spread usage of network and graph based approaches in modeling data has been approved to be effective for various applications. The network based framework becomes more powerful when it is expanded to benefit from the widely available techniques for data mining and machine learning which allow for effective knowledge discovery from the investigated domain. The underlying reason for the substantial efficacy in studying graphs, either directly (i.e., data is given in graph format, for example, the “phone-call” network in studying social evolutions) or indirectly (network is inferred from data by predefined method or scheme, such as co-occurrence network for studying genetic behaviors), is the fact that graph structures emphasize the intrinsic relationship between entities, i.e., nodes (or vertices) in the network (in this chapter, the terms network and graph are used interchangeably). For the indirect case information extraction techniques may be adapted to investigate open sources of data in order to derive the required network structure as reflected in the current available data. This is a tedious process but effective and could lead to more realistic and up-to-date information reflected in the network. The latter network will lead to better and close to real-time knowledge discovery in case online information extraction is affordable and provided. Estimating network structure has attracted the attention of other researchers involved in terrorist network analysis, e.g.[9].