Data Mining Practical Machine Learning Tools and Techniques

We will review some of the major results in random graphs and some of the more challenging open problems. We will cover algorithmic and structural questions. We will touch on newer models, including those related to the WWW.

Random graphs

Twitter, a microblogging service less than three years old, commands more than 41 million users as of July 2009 and is growing fast. Twitter users tweet about any topic within the 140-character limit and follow others to receive their tweets. The goal of this paper is to study the topological characteristics of Twitter and its power as a new medium of information sharing.We have crawled the entire Twitter site and obtained 41.7 million user profiles, 1.47 billion social relations, 4,262 trending topics, and 106 million tweets. In its follower-following topology analysis we have found a non-power-law follower distribution, a short effective diameter, and low reciprocity, which all mark a deviation from known characteristics of human social networks [28]. In order to identify influentials on Twitter, we have ranked users by the number of followers and by PageRank and found two rankings to be similar. Ranking by retweets differs from the previous two rankings, indicating a gap in influence inferred from the number of followers and that from the popularity of one's tweets. We have analyzed the tweets of top trending topics and reported on their temporal behavior and user participation. We have classified the trending topics based on the active period and the tweets and show that the majority (over 85%) of topics are headline news or persistent news in nature. A closer look at retweets reveals that any retweeted tweet is to reach an average of 1,000 users no matter what the number of followers is of the original tweet. Once retweeted, a tweet gets retweeted almost instantly on next hops, signifying fast diffusion of information after the 1st retweet.To the best of our knowledge this work is the first quantitative study on the entire Twittersphere and information diffusion on it.

https://ink.library.smu.edu.sg/cgi/viewcontent.cgi?article=7104&context=sis_research

What is Twitter, a social network or a news media?

The term “Linked Data” refers to a set of best practices for publishing and connecting structured data on the Web. These best practices have been adopted by an increasing number of data providers over the last three years, leading to the creation of a global data space containing billions of assertions— the Web of Data. In this article, the authors present the concept and technical principles of Linked Data, and situate these within the broader context of related technological developments. They describe progress to date in publishing Linked Data on the Web, review applications that have been developed to exploit the Web of Data, and map out a research agenda for the Linked Data community as it moves forward.

/pdf/linked-data-the-story-so-far-2a9zvmaysa.pdf

Linked Data - the story so far

Obesity now affects one in five children in the United States. Discrimination against overweight children begins early in childhood and becomes progressively institutionalized. Because obese children tend to be taller than their nonoverweight peers, they are apt to be viewed as more mature. The inappropriate expectations that result may have an adverse effect on their socialization. Many of the cardiovascular consequences that characterize adult-onset obesity are preceded by abnormalities that begin in childhood. Hyperlipidemia, hypertension, and abnormal glucose tolerance occur with increased frequency in obese children and adolescents. The relationship of cardiovascular risk factors to visceral fat independent of total body fat remains unclear. Sleep apnea, pseudotumor cerebri, and Blount's disease represent major sources of morbidity for which rapid and sustained weight reduction is essential. Although several periods of increased risk appear in childhood, it is not clear whether obesity with onset early in childhood carries a greater risk of adult morbidity and mortality. Obesity is now the most prevalent nutritional disease of children and adolescents in the United States. Although obesity-associated morbidities occur more frequently in adults, significant consequences of obesity as well as the antecedents of adult disease occur in obese children and adolescents. In this review, I consider the adverse effects of obesity in children and adolescents and attempt to outline areas for future research. I refer to obesity as a body mass index greater than the 95th percentile for children of the same age and gender.

Health Consequences of Obesity in Youth: Childhood Predictors of Adult Disease

It is has long been anecdotally known that web archives and search engines favor Western and English-language sites. In this paper we quantitatively explore how well indexed and archived are Arabic language web sites. We began by sampling 15,092 unique URIs from three different website directories: DMOZ (multi-lingual), Raddadi and Star28 (both primarily Arabic language). Using language identification tools we eliminated pages not in the Arabic language (e.g., English language versions of Al-Jazeera sites) and culled the collection to 7,976 definitely Arabic language web pages. We then used these 7,976 pages and crawled the live web and web archives to produce a collection of 300,646 Arabic language pages. We discovered: 1) 46% are not archived and 31% are not indexed by Google (www.google.com), 2) only 14.84% of the URIs had an Arabic country code top-level domain (e.g., .sa) and only 10.53% had a GeoIP in an Arabic country, 3) having either only an Arabic GeoIP or only an Arabic top-level domain appears to negatively impact archiving, 4) most of the archived pages are near the top level of the site and deeper links into the site are not well-archived, 5) the presence in a directory positively impacts indexing and presence in the DMOZ directory, specifically, positively impacts archiving.

/pdf/how-well-are-arabic-websites-archived-cgdjs1ezho.pdf

How Well Are Arabic Websites Archived

Human-generated collections of archived web pages are expensive to create, but provide a critical source of information for researchers studying historical events Hand-selected collections of web pages about events shared by users on social media offer the opportunity for bootstrapping archived collections We investigated if collections generated automatically and semi-automatically from social media sources such as Storify, Reddit, Twitter, and Wikipedia are similar to Archive-It human-generated collections This is a challenging task because it requires comparing collections that may cater to different needs It is also challenging to compare collections since there are many possible measures to use as a baseline for collection comparison: how does one narrow down this list to metrics that reflect if two collections are similar or dissimilar? We identified social media sources that may provide similar collections to Archive-It human-generated collections in two main steps First, we explored the state of the art in collection comparison and defined a suite of seven measures (Collection Characterizing Suite - CCS) to describe the individual collections Second, we calculated the distances between the CCS vectors of Archive-It collections and the CCS vectors of collections generated automatically and semi-automatically from social media sources, to identify social media collections most similar to Archive-It collections The CCS distance comparison was done for three topics: "Ebola Virus," "Hurricane Harvey," and "2016 Pulse Nightclub Shooting" Our results showed that social media sources such as Reddit, Storify, Twitter, and Wikipedia produce collections that are similar to Archive-It collections Consequently, curators may consider extracting URIs from these sources in order to begin or augment collections about various news topics

Bootstrapping Web Archive Collections from Social Media

Visualization eases insight into complex systems such as co-authorship networks We present an initial deployment of an author navigator application for convenient visual examination of JCDL and LANL co--authorship networks

Toolkits for visualizing co-authorship graph

For bounded datasets such as the TREC Web Track (WT10g) the computation of term frequency (TF) and inverse document frequency (IDF) is not difficult. However, when the corpus is the entire web, direct IDF calculation is impossible and values must instead be estimated. Most available datasets provide values for term count (TC) meaning the number of times a certain term occurs in the entire corpus. Intuitively this value is different from document frequency (DF) , the number of documents (e.g., web pages) a certain term occurs in. We investigate the relationship between TC and DF values of terms occurring in the Web as Corpus (WaC) and also the similarity between TC values obtained from the WaC and the Google N-gram dataset. A strong correlation between the two would gives us confidence in using the Google N-grams to estimate accurate IDF values which for example is the foundation to generate well performing lexical signatures based on the TF-IDF scheme. Our results show a very strong correlation between TC and DF within the WaC with Spearman's ρ *** 0.8 (p ≤ 2.2×10*** 16) and a high similarity between TC values from the WaC and the Google N-grams.

/pdf/correlation-of-term-count-and-document-frequency-for-google-8nrjb1w6bx.pdf

Correlation of Term Count and Document Frequency for Google N-Grams

The Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH) began as an alternative to distributed searching of scholarly e-print repositories. The model embraced by the OAI-PMH is that of metadata harvesting, where value-added services (by a "service provider") are constructed on cached copies of the metadata extracted from the repositories of the harvester's choosing. While this model dispenses with the well-known problems of distributed searching, it introduces the problem of synchronization. Stated simply, this problem arises when the service provider's copy of the metadata does not match the metadata currently at the constituent repositories. We define some metrics for describing the synchronization problem in the OAI-PMH. Based on these metrics, we study the synchronization problem of the OAI-PMH framework and propose several approaches for harvesters to implement better synchronization. In particular, if a repository knows its update frequency, it can publish it in an OAI-PMH identify response using an optional about container that borrows from RDF Site Syndication (RSS) Format.

/pdf/repository-synchronization-in-the-oai-framework-37ef8el6l2.pdf

Michael L. Nelson

Papers

How Well Are Arabic Websites Archived

Bootstrapping Web Archive Collections from Social Media

Toolkits for visualizing co-authorship graph

Correlation of Term Count and Document Frequency for Google N-Grams

Repository synchronization in the OAI framework