I was actually very interested in the Conferences project that you explained to the class and I decided to make a working prototype so that I can work on personally.
I gained access to a Linux based server with MySQL and I started to make a small demo using php. I have the database set up and I put in one demo entry and I created a page that should more or less view what is in the database. I unfortunately don't have access to the error logs of the server and I have errors. I am currently waiting on the server administrator to create the script that I wrote to send my errors to my personal folder.
So far I have created a page that should (emphasis on the should) connect to the database and display the entries there, a page to add entries, the database connection, and a common file that takes care of most of the repetitive functions that the site will call.
I realized today why the pages were not displaying properly. The server has not been configured properly translate the php and the MySQL connections. I have purchased a hard drive and will be installing Fedora Core 7 and setting up my server this weekend.
I have attempted to install the new hard drive to no avail. The motherboard won't recognize it and I have had to resort to unconventional means to get my server running again. I have reinstalled Linux, Apache, MySQL, PHP (LAMP)combination and set my DNS and Port Forwarding on my router to point to the server.
I had an old copy of a website that I had to test see if I had it configured properly and I still had a problem viewing my web page within my domain much less outside of my domain. I had apparently set the http configuration were set to listen to the wrong port (Port 40) and my DNS settings were set to the wrong port.
I now know that my server is configured correctly to ordinary HTML pages but I am still unsure as to whether I will have access to the the database using PHP with my current configurations and as such it is and that will be the next project.
When I have this server configured properly with access to the database, I hope to have a small demo of simple program to get more familiarized with LAMP. I also plan to do research on current database applications as it pertains calendars.
I have been reading the chapter in the class text, Beautiful Code, called "When a Button is All that Connects You to the World". I found it interesting because it combined the use of software and hardware to create something helpful to the handicapped and it reminded me of research I did with the Sony AIBO and Tekkotsu.
Sony created the AIBO as an electronic pet for professionals. They were not as successful as they had hoped but they provided relatively cheap hardware for some students studying Cognitive Robotics especially Ethan J. Tira-Thompson who created Tekkotsu as software that would sit on top of Sony's proprietary software and manipulate the movements, actions and decision making algorithms of the AIBO.
After talking to Dr. Abbott after class last week, he directed to do some reading in Natural Language processing. I found an interesting site on this topic from a Research Center in Palo Alto.
I read the paper Precision-focused Textual Inference and I will be presenting a short overview of this paper.
I have been doing more reading with regards to Natural Language Processing with more of an emphasis on Information Retrieval. The major subtasks of this process are:
- Named Entity Recognition: recognition of entity names (for people and organizations), place names, temporal expressions, and certain types of numerical expressions.
- Coreference: identification chains of noun phrases that refer to the same object. For example, anaphora is a type of coreference.
- Terminology extraction: finding the relevant terms for a given body of text
I've decided to tackle each topic one at a time. This week I have done some light reading on Named Entity Recognition and I have been demoing some free Information Retrieval Software.
An example can be found here.
As I mentioned last week I have moved to reading on Coreference. I found an interesting paper on text extraction where the text first undergoes the Name Recognition using software called Sylex(last week's topic), then undergoes Term Extraction using the software, LEXTER (this week's topic), and finally Term Clustering (next week's topic) which uses FASTR.
I have also been working on bringing a working demo online and so far I have gotten just the framework together. I now have the web server running and it recognizes PHP. Now I'm working on getting the database up and running so I can post and retrieve data. I am currently working on the aesthetics and it's coming along nicely.
As mentioned in my previous entries I finished reading the third approach of Mamed Recognition, Terminology Extraction. I have begun to formulate an prospectus while researching development of UML diagrams to describe the User Interface and the different components of the Conference system. I have also looked into incorporating additional features along with the consideration of the Operating System, Database, nad Existing systems that I might want to use in the final project.
I have been coding the actually system to view, add, update and delete entries. I chose to use the Linux Environment, specifically Fedora Core 8 with a MySQL Database. I have set up an Apache web server that supports PHP which is the language I will be writing most of the code. I downloaded Lucene which is written in Java and although I am having some issues with Lucene's requirements and it's compatibility of the current version of Fedora Core I hope to get it working by the end of the week.
This past week has been a matter of trial and error when it comes to System Administration (something that I have always been an end user of). I am painstakingly learning the more subtle intricacies of PHP and MySQL relatively on my own.
I've also run through the outline of a Prospectus and I am still toying with the idea of actually doing a Master's project.
I started on my first draft of my Prospectus this week. I put the Lucene project on hold till late this week. After speaking with Dr. Abbot last week. I have an idea of the outline of a project that I can start writing a prospectus step by step.
I was supposed to have presented a start of my prospectus this week but after speaking with Dr. Abbott last week we came to a consensus that I would just consider the Conference Application as a project for the class and concentrate on CoolStateLA.
I had put aside any coding on this application and as such all the updates that I should have been working on are incomplete. I worked out and implemented the algorithm for ensuring that there isn't a conflict with regards to dates.
I have written the cron job for updating the call for papers page so that call for papers with past deadlines will not be displayed. I hopefully will have the ability to add users that will have the ability to choose the conferences they are interested in and display them. They will only be users who cannot edit the data. The only user who will be able to edit, delete or add new conference/call for papers for now will bet he superuser/admin.
I have given up trying to get an external search feature to work and concentrate on getting the actual application to work use MySQL search feature for any searching. I have created a login and sign up page for users to access the conferences and search for conferences they want to view. To facilitate that I had to created a new table called user.
There is a new field on most of the pages called category and I had to create 2 new tables to store the data I have gathered. Category: A list of categories that a conference might fall under. Categorize: A list of all the conference id from the conference table and their corresponding categories
I have given the user the option to sort by , country, state, between a range of call for paper deadline dates, and a range of conference start dates, and categories. To get a list of all the conferences they simply have to leave all the data blank.
Known Bugs/Future Additions
Implementation of sessions
Security to admin
Search for Conference Start correction
Delete functionality correction