Hello people, I stumbled upon this forum and I see very positive and intelligent insights being given by the community. =)

I am currently in my final year and i have this idea of using Natural Language Processing techniques to assist the development of my project.

I plan to build a Personal Scheduler that will be able to understand human input by using Chat Bot methodology and successfully organize tasks/appointments according to their level of priority.

Regards,
Kevin

You might get some ideas from “Remember the Milk”

http://www.rememberthemilk.com/

Hi, Kevin, and welcome to chatbots.org. We have a lot of threads which relate to a project of that sort, and a great number of other interesting discussions here, as well. Feel free to look around, and if you have any questions, don’t hesitate to ask.

Welcome to the forum, Kevin!

Your project reminds me of food-bot, a program developed to search college emails for events and strip out the time, date, organizers, etc. from the email. The bot assesses whether there will be free food at the event, what that food is, and even how “awkward” it would be to show up!

You would probably want your bot to operate in a similar manner, stripping time, date, and other important details from conversation and loading that data into an organizer (calendar, etc). You’d have to be able to do a broader search than food-bot in order to determine the nature of the event. Can’t just search for food-related words, or words related to any such narrow topic. This seems to me like the biggest challenge.

At any rate, sounds like a cool project. What tools were you thinking of using to implement the chatbot portion of the program?

Hello, thanks for the replies. =)

I forgot to mention that I will be using Java for the project. It will be calender based, and the project will allow the user to select either tasks/appointments. When an input is entered, the system will scan for keywords from the sentence and help prioritize tasks/appointments according to their level of importance (Color).

I have done some research and found out this method which is called Recursive Distributed Representations, which will be using a modified form of a backprop neu network to help solve the syntactical indifference of most NLP projects.

For word sense confusion however, I might need the help of topic model API’s as the project scope might be too wide.

The project uses Chat bot methodology in the sense that it does not formulate a response for the user, it will just take the input, and assign a weight(integer) to each representation so that when two or more tasks/appointments come clashing, they can be prioritized accordingly.

Since i will be implementing the system in java. is there any recommended tools related to NLP that i can use? Sorry for the double post =)

Here’s a link to a java implementation of WordNet. I’ve never tried it, so if you decide to use it, please post what you think.

I also has the problem with word sense and currently my design is identify nouns of user entered sentence using a pos tagger and replace them with the synonyms that taken from word net. This is not good method since word net some times retuning large number of synonyms.

I found another sullution here( http://blog.opencog.org/2009/01/12/determining-word-senses-from-grammatical-usage/) with Relex frame work. It is good one but according to it take time.

If you find a good solution to identify synonyms(word sense) and other natural language problems share it.

Charith Soori - Mar 30, 2011:

I also had the problem with word sense and currently my design is to identify nouns of user entered sentence using a pos tagger and replace them with the synonyms that taken from word net. This is not good method since word net some times returning large number of synonyms.

I found a another solution here( http://blog.opencog.org/2009/01/12/determining-word-senses-from-grammatical-usage/) that used in Relex frame work. It is good one but according to it take time.May be they there will be a way to reuse there work since it is a open source project. But I didn’t try.(If some one have experience ..)

If you find a good solution to identify synonyms(word sense) and other natural language problems share it.

Thank you, but what do most people do when it comes to determining the syntactical structure of the sentence?
I have read about Pollack’s Recursive Distributed Representations but am not sure how to implement it. Seems like it makes use of a parse tree, a dictionary and a Recursive auto-associative memory. Very confusing stuff. =(

What is the method you using for chatbot.(i.e alice ,eliza ,pue nlp) ?

i will be replacing the keyword scanning mechanism used by ELIZA with recursive distributed representations. Still doing research though.

Some time dis research help you to (A Comparison Between Alice and Elizabeth Chatbot Systems)

Okay thanks, have u seen Jabberwacky too? I cant find any sample implementations that relates to jabberwacky anywhere =(

Ya I am too interested it. According what I found It has good learning module (About the Jabberwacky AI). That is the reason I want to follow it. I am searching good learning module other than aiml <learn>. If you found anything send a notification . :D

but in what language are u programming in?