CA1

This post is for grading purposes. It's essentially a summary of the work I have done thus far, as well as an indication of future goals.

Summary of Progress and Results
Literature Review:
Read the following papers:
1) Poetry Mix-up: Preserving Culture with Poetic SMS by Adrian David Cheok, Owen Noel Newton Fernando, Janaka Prasad Wijesena, and Ramkumar
Shankar
2) BlogWall: Displaying Artistic and Poetic Messages on Public Displays via SMS by Adrian David Cheok, Abd-ur-Rehman Mustafa, Owen Noel Newton Fernando, Anne-Katrin Barthoff, Janaka Prasad Wijesena, Naoko Tosa
3) BlogWall: Personal and Poetic Expressions by Owen Noel Newton Fernando, Janaka Prasad Wijesena, Adrian David Cheok, Wei Liu, and Naoko Tosa
4) NLUS - a Prolog Based Natural Language Understanding System by M.C. Lee, and H.V. Leong
5) The Research on the Application of Text Clustering and Natural Language Understanding in Automatic Abstracting by Qinglin Guo and Cunbin Li
6) The Use of Reasoning in Natural Language Understanding by Derek Long and Roberto Garigliano
7) Bidirectional Inference with the Easiest-First Strategy for Tagging Sequence Data by Yoshimasa Tsuruoka and Jun’ichi Tsujii

Also did research/reading up on the Internet, about Natural Language Understanding and Sentence Parsing in general.
Books I am still reading:
1) Natural Language Processing and Knowledge Representation - Language for Knowledge and Knowledge for Language edited by Łucja M. Iwańska and Stuart C. Shapiroa
2) Artificial intelligence : structures and strategies for complex problem solving by George F. Luger
3) Handbook of natural language processing edited by Robert Dale, Hermann Moisl, Harold Somers

Other Work:
1) Familiarised myself (and experimented) with the workings of the Blogwall system.
2) Installed the Blogwall system and MySQL Database on my own system.
3) Prepared for Blogwall presentation to visitors from Swedish Institute of Technology (although we didn't get to present due to lack of time).
4) Familiarised and experimented with the Magic Classes in the code.
5) Installed and ran tests on Makino's latest visualisation code (to test for memory leaks).
6) Extracted definitions from the dictionary for each of the chosen keywords from the SMS input.
7) Carried out PosTagging on each of the definitions extracted, for further analysis and understanding.
8) Attempting to isolate cause of memory leak problems (yet to be completed).

Future Plans/Schedule
1) Expedite resolution of memory leak problems in visualisation code (the cause of this is most probably in the polling portion of the code).
2) Use the PosTagged definitions of the chosen keywords to provide better understanding of the meaning of these keywords (co-relating the object in the SMS input to the object in the definition?).
3) Utilise the rest of the text in the SMS that are not chosen, to provide better understanding of the context of the SMS in its entirety (Perhaps the dictionary may prove useful here again?).

These provide possible next steps that I can take, but as of now, I am still unclear as to where they will lead me, in terms of SMS input analysis. I believe I will get a clearer picture and more ideas after reading through the books I have on hand (rather than the papers, since they do not provide any background on the topic of Natural Language Understanding but delve straight into their areas of study and as such, have not proven to be very comprehensive -- which is why Ken suggested to myself to start with the books).

I intend to isolate the memory leak problem early next week (Monday) and start debugging it. Hopefully it will be resolved by the end of next week. If I manage to resolve it early in the week, then I can work on finding a way to utilise the PosTagged definitions.
Comments