Você está na página 1de 6

Rsum

Name:

John Conroy
Bio & Contact

Age: Address:

33 46 Cloonarkin Drive, Oranmore, Co. Galway j.conroy3@nuigalway.ie conroyform01@yahoo.com 087 7565863 Ph.D. candidate (since Sep. 2010) in computer science/software development, specialising in computational complex-graph dynamics, graph theory, computational social network analysis, and information retrieval.

Email: Phone: Current employment status:

Overview Highly skilled Python and Java programmer strong computing and mathematical fundamentals a special interest in Web technologies Special interest in graphs/networks theory. High aptitude for communication and for creativity. I hold a M.Sc. (a 1:1 @77%) in software development Significant practical coding experience as a freelancer. I have been coding since 2008, when I changed careers and returned to university to pursue a Masters in software development. I am thus looking for my first full-time job in the industry.

j.conroy3@nuigalway.ie

0877565863

Relevant skills

Programming Languages:

Strongest: Intermediate: Weaker

Java, Python, PHP, HTML, SQL Javascript, CSS, C#, ASP.NET C++,VB,C

Mathematical/Data analysis: Statistics: hypothesis/significance testing, regression/correlation, data analysis Linear algebra/vector-space modelling Machine learning, information retrieval/natural language processing Complex graph theory and common related algorithms/operations (cuts, cliques, network robustness, centrality, weighted-graph operations etc.) Generally high mathematical competence

Web APIs
Twitter, facebook, LinkedIn Tools/Frameworks/databases, formats, other: Eclipse, Linux, Windows, Django, MySQL, Web spiders/crawlers, Minitab, XML and JSON, Wordpress, Drupal, basic Apache, basic R

Programming Projects/experience
In Java

rsTwitterClient
A twitter desktop client written in Java. A custom, business intelligence-oriented desktop twitter client for a SME customer in Galway (contract work, also required several standalone scripts for various batch processing tasks).

Technology:
Java SE GUI (Swing and AWT) Multi-threading Interfaces with the twitter Search and REST apis. External libraries: jtwitter (twitter api wrapper), jxl(create/read/manipulate MS Excel files in Java)

Features:
Enhanced desktop twitter search engine for marketing/business intelligence

j.conroy3@nuigalway.ie

0877565863

Convenient tabulated results with user metadata (#followers, #listed, location, geolocation etc.) Sort results by user metadata, date, whatever Follow button for users Post messages from client One-click add multiple users to message Quick-message capabilities, field for hyperlinks you most commonly share Save all results + all details on each user in search results Streaming mode: auto-search keyword every n minutes, saves to .xls file Intelligent streaming/saving: duplicates removed Multiple twitter accounts enabled from one interface

In Python

I primarily use Python for my research work. I use Python as my primary language for most tasks for its ease/speed of development
For my research (M.Sc. and Ph.D.): Utilising the twitter API and Bit.ly API to collect a large dataset (millions of items) of twitter data and to decode URLs Screen-scraping several thousand pages from YouTube, extracting important tokens like title, video tags, category etc. Analysing/aggregating this data. Building a social graph from connections between twitter users. Building a second-order social graph based on conversations between users. Analysing these graphs in numerous ways including identifying important users using a modified PageRank-type algorithm, and comparing this with other link-analysis methods Developing and implementing various algorithms for network traversal Developing and implementing a new algorithm for weighting activity graphs in social networks Identifying key network metrics such as cliques, Pearsons R, min-cuts etc. Implementing two supervised machine-learning algorithms to train a classifier to classify tweets in terms of sentiment (was a tweet positive or negative in sentiment towards a brand?). Nave Bayes and Maximum Entropy Vector-space analysis of text documents, writing scripts for tf-idf weighting and cosinesimilarity measurement of documents. Identifying and implementing various preprocessing/storage/analysis solutions for same, too numerous to mention Efficient implementation of above scripts for large volumes of data (up to gigabytes) A very few examples available at github: https://gist.github.com/johnconroy Graphing with MatPlotLib and R, statistical analysis of data. Web development in Python: Im intermittently working on a web application written in Python using Django (a RAD Web development framework). This application enables on-the-fly generation of news hubs for breaking news topics with smart filtering/scraping technologies. This service is in the early stages of development. Social network data analysis & business intelligence: I performed some work recently for a start-up Web-publishing company which required: Extracting user data for the company twitter follower group, facebook page, and LinkedIn user page group Aggregating this data in several ways for reporting on user demographics etc. NLP on this data for key terms in descriptions, job titles etc., interfacing with a graphics library for visualisation of key terms Interfacing with LinkedIn Search API for further info on these users
j.conroy3@nuigalway.ie 0877565863

I also wrote scripts for this company to: Identify trending topics in their field, using rival publisherss headlines/articles, applying information-retrieval concepts such as Zipfs distribution plus various heuristics

In PHP http://www.footiechin.com Imageboard/4-chan clone written in PHP. I never marketed/launched it: I just wanted a project to learn PHP (only one board is working, requires trivial but time-consuming work to launch other boards, fix minor bugs etc.). Features are all built from the ground up in pure php no plugins/frameworks etc. Incorporates user management with different levels of access, a custom Capcha system, anonymous or registered user posting, robust user session management, threads/posts, image uploads, random image assignment when no image is uploaded by the user, and other features common on similar boards on the Web. Technical details (& architecture) available in blog posts linked to at the bottom of the home page. In other languages

C# & ASP.NET
ACME Warehouse Management System: For the first year of my Masters degree, two teammates and I were required to implement a practical project. Our team built a Web-based logistics (warehouse) management system in C# + ASP.net + Javascript + SQL Server, for which we won best project for the year http://bit.ly/bwiIbi . The system incorporated a GUI warehouse management storage system, granular user management, various reporting and automation features and other features.

Other Web development:


Ive built a number of web sites for myself and others using CMSs (WordPress and Drupal), and basic HTML/CSS/JavaScript going back to ~2000.

Education:

M.Sc. Software Development (NUI Galway), 2010 First class honours: 77% Year 1- classroom based: 72%

C, C++, VB.net, RDBMS, Web Programming (HTML,C#,Javascript) Group project, Software Engineering methods.

Year 2- thesis: 83% Social Network Analysis and Sentiment Analysis of Microblogging Data:
Aquire datasets from Twitter social network using their APIs, isolated and analysed the network graph of Irish users for key metrics Eg o implement PageRank algorithm to determine authoritative users, investigate hybrid link-analysis methods to determine same o Analyse network according to edge centrality, cuts, network topology, scaling coefficient etc.
0877565863

j.conroy3@nuigalway.ie

Sentiment analysis: collect a large dataset of tweets, use machine learning methods (Nave Bayes and Maximum Entropy models) to train & build a sentiment classifier to assess twitter users response to TV commercials during the Superbowl. I presented a paper on this work at the AICS conference in August 10

Ph.D. (ongoing): Complex dynamic networks, social networks, and information retrieval
Recently completed work involved data mining of social network data, and using IR techniques (vector space, tf-idf and cosine-similarity) to compare users of a social network. I then compared user similarity as a function of connectedness in various social graphs (implied and formal social graphs) in which they reside. This novel work represents an empirical measurement of similarity between social network users based on what they have written and thus provided a platform for empirical evaluation of users as a function of how strong the social connections between them are. Ongoing work is in two areas. First, an innovative weighting scheme (a formula) for the strength of social ties between users in social networks, which will hopefully be useful in recommender systems. Second, a complex-graph generation model based on the classical preferential attachment approach, but augmented by simulations of how users might link to each other based on similarity between them. I then analyse the graphs created by this augmented model to see if they better represent growth of real complex networks (social networks, biological networks, telecommunications networks etc.)

Other education: B.Comm, (NUI Galway) 2001 2:2, minors in Information Systems and Economics Leaving Cert 96 465 points, A2 in English(hons), B1 in Maths (hons)

Other abilities/skills Academic:


Relief lecturing in Programming (Python) and a Computer Applications module (Internet/Web history & architecture). Lab assistant for programming (Python) and Databases (a post-grad data-modelling and RDBMS theory & practise class). Review/catchup classes to undergrads in Python and Internet programming. Grinds in programming (Java and Python), algotithms, databases. I have acted as exam invigilator and corrector of exams in the department.

Written, verbal communication:


Writing was my hobby for many years, although I dont have much time for it any more. As well as amateur fiction, I used to do a lot of tech. blogging for a Silicon Valley trade ezine: cmswire.com. Ive written comedic articles for The Dubliner, and have also contributed pieces for The Mail on Sunday (a book review) and The Village (a piece on energy supply). As well as tutoring and occasional relief lecturing, I spoke at two conferences in 2010 and one in 2011. I presented a paper at the AICS conference, and also spoke on the subject of Twitter for business and researchers at the BlogTalk conference at NUI Galway. A video of the latter is available here: http://vimeo.com/14797222
j.conroy3@nuigalway.ie 0877565863

Referees: Colm ORiordan Lecturer, Information Technology, National University of Ireland, Galway. Ph. 091 49366 Josephine Griffith Lecturer, Information Technology, National University of Ireland, Galway. Ph 091493717 John Breslin Co-founder Boards.ie, founder NewTechPost, founder Streamglider, lecturer, author: http://www.johnbreslin.com/contact Gavin Duffy Managing Director, RealSim, http://www.realsim.ie , 091 444127

j.conroy3@nuigalway.ie

0877565863

Você também pode gostar