You can read a nice introduction to various techniques used to design chatbots here: http://www.gamasutra.com/view/feature/6305/beyond_fa%C3%A7ade_pattern_matching_.php

Also, here are a few useful links:

 - http://web.archive.org/web/20120320060043/
 - http://ai-programming.com/bot_tutorial.htm
 - http://www.alicebot.org/be.html
 - http://en.wikipedia.org/wiki/List_of_chatterbots
 - http://www.codeproject.com/Articles/36106/Chatbot-Tutorial
 - http://www.slideshare.net/amyiris/ai-and-python-developing-a-conversational-interface-using-python

The [Natural Language Toolkit (python)][1] implements a few chatbots: http://nltk.github.com/api/nltk.chat.html

Simple pipeline architecture for a spoken dialogue system from the book [Natural Language Processing with Python - Analyzing Text with the Natural Language Toolkit][2] By&#160;Steven Bird, Ewan Klein, Edward Loper:

[![enter image description here][3]][3]


  [1]: http://www.nltk.org
  [2]: http://shop.oreilly.com/product/9780596516499.do
  [3]: http://i.stack.imgur.com/aivDD.png

The two places I would start with are [how cleverbot works [part of a podcast]](http://www.radiolab.org/2011/may/31/) and then go through the [Natural Language Toolkit Book](http://www.nltk.org/book) to learn about the algorithms to use. (NLTK uses python, but the book is also a python tutorial)

I have a scrolling element on my page (with the jScrollPane jQuery plugin). What I want to accomplish is a way to turn off the scrolling window by detecting the width of the browser window. I am doing a responsive layout and I want this scrolling feature to be turned off when the browser is below a certain width. I am able to make it work when I refresh the page, but when I resize the browser window the width value does not update on the fly. 

Right now if I start out with a window that is 1000px wide then resize to 350px the scroll feature remains. I want the scroll feature to shut off as soon as the browser width hits 440px.

Here&#39;s the code I have so far..

    var windowsize = $(window).width();

    $(window).resize(function() {
      var windowsize = $(window).width();
    });
  
    if (windowsize &gt; 440) {
      //if the window is greater than 440px wide then turn on jScrollPane..
        $(&#39;#pane1&#39;).jScrollPane({
	       scrollbarWidth:15, 
	       scrollbarMargin:52
        });
    }

jQuery: How to detect window width on the fly?

I have a REST URL to get all users formatted like this:
    http://example.com/users

To get an individual user by id:
    http://example.com/users/12345

To get all user&#39;s bids:
    http://example.com/users/12345/bids

To get all user&#39;s bids between two dates:
    http://example.com/users/12345/bids/?start=01/01/2012&amp;end=01/31/2012

or should it be like this:
    http://example.com/users/12345/bids/start/01012012/end/01312012

I&#39;m leaning towards the 1st date range URL as start and end are not entities in the domain. What is the proper way to format a REST URL with a date range?

Thanks,

Tom

Proper REST formatted URL with date ranges

As a engineering student, I would like to make a chat bot using python. So, I searched a lot but couldn&#39;t really find stuff that would teach me or give me some concrete information to build a intelligent chat bot.

I would like to make a chatbot that gives human-like responses (Simply like a friend chatting with you). I am currently expecting it to be as just a software on my laptop (would like to implement in IM, IRC or websites later).

So, I am looking for a tutorial/ any other information which would certainly help me to get my project done.



Any tutorials for developing chatbots?

As a engineering student, I would like to make a chat bot using python. So, I searched a lot but couldn't really find stuff that would teach me or give me some concrete information to build a intelligent chat bot.
I would like to make a chatbot that gives human-like responses (Simply like a friend chatting with you). I am currently expecting it to be as just a software on my laptop (would like to implement in IM, IRC or websites later).
So, I am looking for a tutorial/ any other information which would certainly help me to get my project done.

I know that if I change the linewidth of a line, that is automatically updated in the legend.
However I would like to just change the legend linewidth without affecting the plot.

increase the linewidth of the legend lines in matplotlib

I have a simple image that I&#39;m showing with imshow in matplotlib. I&#39;d like to apply a custom colormap so that values between 0-5 are white, 5-10 are red (very simple colors), etc.  I&#39;ve tried following this tutorial: 

http://assorted-experience.blogspot.com/2007/07/custom-colormaps.html with the following code:



    cdict = {
    &#39;red&#39;  :  ((0., 0., 0.), (0.5, 0.25, 0.25), (1., 1., 1.)),
    &#39;green&#39;:  ((0., 1., 1.), (0.7, 0.0, 0.5), (1., 1., 1.)),
    &#39;blue&#39; :  ((0., 1., 1.), (0.5, 0.0, 0.0), (1., 1., 1.))
    }
    
    my_cmap = mpl.colors.LinearSegmentedColormap(&#39;my_colormap&#39;, cdict, 3)
    
    plt.imshow(num_stars, extent=(min(x), max(x), min(y), max(y)), cmap=my_cmap)
    plt.show()
    
But this ends up showing strange colors, and I only need 3-4 colors that I want to define.  How do I do this?

Defining a discrete colormap for imshow in matplotlib

I&#39;m having a hard time wrapping my brain around [PEP 380][1].

 1. What are the situations where `yield from` is useful? 
 2. What is the classic use case?
 3. Why is it compared to micro-threads?

So far I have used generators, but never really used coroutines (introduced by [PEP-342][2]). Despite some similarities, generators and coroutines are basically two different concepts. Understanding coroutines (not only generators) is the key to understanding the new syntax.

IMHO **coroutines are the most obscure Python feature**, most books make it look useless and uninteresting.

---

Thanks for the great answers, but special thanks to [agf][3] and his comment linking to [David Beazley presentations][4].

  [1]: http://www.python.org/dev/peps/pep-0380/
  [2]: http://www.python.org/dev/peps/pep-0342/
  [3]: https://stackoverflow.com/users/500584/agf
  [4]: http://www.dabeaz.com/coroutines/

In practice, what are the main uses for the &quot;yield from&quot; syntax in Python 3.3?

I use an external module ([libsvm][1]), which does not support numpy arrays, only tuples, lists and dicts. But my data is in a 2d numpy array. How can I convert it the pythonic way, aka without loops. 


    &gt;&gt;&gt; import numpy
    &gt;&gt;&gt; array = numpy.ones((2,4))
    &gt;&gt;&gt; data_list = list(array)
    &gt;&gt;&gt; data_list
    [array([ 1.,  1.,  1.,  1.]), array([ 1.,  1.,  1.,  1.])]

    &gt;&gt;&gt; type(data_list[0])
    &lt;type &#39;numpy.ndarray&#39;&gt;  # &lt;= what I don&#39;t want

    # non pythonic way using for loop
    &gt;&gt;&gt; newdata=list()
    &gt;&gt;&gt; for line in data_list:
    ...     line = list(line)
    ...     newdata.append(line)
    &gt;&gt;&gt; type(newdata[0])
    &lt;type &#39;list&#39;&gt;  # &lt;= what I want


  [1]: http://www.csie.ntu.edu.tw/~cjlin/libsvm/

Convert 2d numpy array into list of lists

I am trying to get the date of the previous month with python.
Here is what i&#39;ve tried:

    str( time.strftime(&#39;%Y&#39;) ) + str( int(time.strftime(&#39;%m&#39;))-1 )

However, this way is bad for 2 reasons: First it returns 20122 for the February of 2012 (instead of 201202) and secondly it will return 0 instead of 12 on January.

I have solved this trouble in bash with

    echo $(date -d&quot;3 month ago&quot; &quot;+%G%m%d&quot;)

I think that if bash has a built-in way for this purpose, then python, much more equipped, should provide something better than forcing writing one&#39;s own script to achieve this goal. Of course i could do something like:

    if int(time.strftime(&#39;%m&#39;)) == 1:
        return &#39;12&#39;
    else:
        if int(time.strftime(&#39;%m&#39;)) &lt; 10:
            return &#39;0&#39;+str(time.strftime(&#39;%m&#39;)-1)
        else:
            return str(time.strftime(&#39;%m&#39;) -1)

I have not tested this code and i don&#39;t want to use it anyway (unless I can&#39;t find any other way:/)

Thanks for your help!


python date of the previous month

I have a large set of vectors in 3 dimensions. I need to cluster these based on Euclidean distance such that all the vectors in any particular cluster have a Euclidean distance between each other less than a threshold &quot;T&quot;.

I do not know how many clusters exist. At the end, there may be individual vectors existing that are not part of any cluster because its euclidean distance is not less than &quot;T&quot; with any of the vectors in the space.

What existing algorithms / approach should be used here?

Unsupervised clustering with unknown number of clusters

What is the difference between *graph search* and *tree search* versions regarding DFS, A* searches in _artificial intelligence_?

What is the difference between graph search and tree search?

I am looking for an open source neural network library.  So far, I have looked at FANN, WEKA, and OpenNN.  Are the others that I should look at?  The criteria, of course, is documentation, examples, and ease of use.  

Open Source Neural Network Library

I was wondering what&#39;s the difference between _uniform-cost search_ and _Dijkstra&#39;s algorithm_. They seem to be the same algorithm.

What&#39;s the difference between uniform-cost search and Dijkstra&#39;s algorithm?

I am trying to build a neural network from scratch.
Across all AI literature there is a consensus that weights should be initialized to random numbers in order for the network to converge faster.

**But why are neural networks initial weights initialized as random numbers?** 

I had read somewhere that this is done to &quot;break the symmetry&quot; and this makes the neural network learn faster. How does breaking the symmetry make it learn faster?

Wouldn&#39;t initializing the weights to 0 be a better idea? That way the weights would be able to find their values (whether positive or negative) faster?

Is there some other underlying philosophy behind randomizing the weights apart from hoping that they would be near their optimum values when initialized?

Why should weights of Neural Networks be initialized to random numbers?

How to check whether a sentence is valid in Python?

Examples:

    I love Stackoverflow - Correct
    I Stackoverflow love - Incorrect

How to check whether a sentence is correct (simple grammar check in Python)?

What I am striving to complete is a program which reads in a file and will compare each sentence according to the original sentence. The sentence which is a perfect match to the original will receive a score of 1 and a sentence which is the total opposite will receive a 0. All other fuzzy sentences will receive a grade in between 1 and 0. 

I am unsure which operation to use to allow me to complete this in Python 3. 

I have included the sample text in which the Text 1 is the original and the other preceding strings are the comparisons.  

Text: Sample 
-
Text 1: It was a dark and stormy night. I was all alone sitting on a red chair. I was not completely alone as I had three cats.

Text 20: It was a murky and stormy night. I was all alone sitting on a crimson chair. I was not completely alone as I had three felines
// Should score high point but not 1

Text 21: It was a murky and tempestuous night. I was all alone sitting on a crimson cathedra. I was not completely alone as I had three felines
// Should score lower than text 20

Text 22: I was all alone sitting on a crimson cathedra. I was not completely alone as I had three felines. It was a murky and tempestuous night.
// Should score lower than text 21 but NOT 0

Text 24: It was a dark and stormy night. I was not alone. I was not sitting on a red chair. I had three cats.
// Should score a 0!


Fuzzy String Comparison

What is the difference between a _constituency parser_ and a _dependency parser_? What are the different usages of the two?

Difference between constituency parser and dependency parser

I&#39;m Working on document classification tasks in java.

 Both algorithms came highly recommended, what are the benefits and disadvantages of each and which is more commonly used in the literature for Natural Language Processing tasks? 



What are the major differences and benefits of Porter and Lancaster Stemming algorithms?

I have some problems to create a new model for Stanford Parser.

I have also downloaded the last version from Stanford: 
http://nlp.stanford.edu/software/lex-parser.shtml

And here, Genia Corpus in 2 formats, xml and ptb (Penn Treebank).

Standford Parser can train with ptd files ; then I downloaded Genia Corpus, because I want to work with biomedical text:

&lt;s&gt;http://categorizer.tmit.bme.hu/~illes/genia_ptb/&lt;/s&gt; &lt;sup&gt;(link no longer available)&lt;/sup&gt;  (genia_ptb.tar.gz)

Then, I have a short Main class to get dependency representation of one biomedical sentence:

        String treebankPath = &quot;/stanford-parser-2012-05-22/genia_ptb/GENIA_treebank_v1/ptb&quot;;
        
        Treebank tr = op.tlpParams.diskTreebank();
        tr.loadPath(treebankPath);	
        LexicalizedParser lpc=LexicalizedParser.trainFromTreebank(tr,op);

I have tried different ways, but always get the same result.

I have an error in the last line. This is my output:

    Currently Fri Jun 01 15:02:57 CEST 2012
    Options parameters:
    useUnknownWordSignatures 2
    smoothInUnknownsThreshold 100
    smartMutation false
    useUnicodeType false
    unknownSuffixSize 1
    unknownPrefixSize 1
    flexiTag true
    useSignatureForKnownSmoothing false
    parserParams edu.stanford.nlp.parser.lexparser.EnglishTreebankParserParams
    forceCNF false
    doPCFG true
    doDep false
    freeDependencies false
    directional true
    genStop true
    distance true
    coarseDistance false
    dcTags false
    nPrune false
    Train parameters: smooth=false PA=true GPA=false selSplit=true (400.0; deleting [VP^SQ, VP^VP, VP^SINV, VP^NP]) mUnary=1 mUnaryTags=false sPPT=false tagPA=true tagSelSplit=false (0.0) rightRec=true leftRec=false collinsPunc=false markov=true mOrd=2 hSelSplit=true (10) compactGrammar=3 postPA=false postGPA=false selPSplit=false (0.0) tagSelPSplit=false (0.0) postSplitWithBase=false fractionBeforeUnseenCounting=0.5 openClassTypesThreshold=50 preTransformer=null taggedFiles=null
    Using EnglishTreebankParserParams splitIN=4 sPercent=true sNNP=0 sQuotes=false sSFP=false rbGPA=false j#=false jJJ=false jNounTags=false sPPJJ=false sTRJJ=false sJJCOMP=false sMoreLess=false unaryDT=true unaryRB=true unaryPRP=false reflPRP=false unaryIN=false sCC=1 sNT=false sRB=false sAux=2 vpSubCat=false mDTV=2 sVP=3 sVPNPAgr=false sSTag=0 mVP=false sNP%=0 sNPPRP=false dominatesV=1 dominatesI=false dominatesC=false mCC=0 sSGapped=4 numNP=false sPoss=1 baseNP=1 sNPNNP=0 sTMP=1 sNPADV=1 cTags=true rightPhrasal=false gpaRootVP=false splitSbar=0 mPPTOiIN=0
    Binarizing trees...done. Time elapsed: 141 ms
    Extracting PCFG...done. Time elapsed: 56 ms
    Compiling grammar...done Time elapsed: 1 ms
    Extracting Lexicon...Exception in thread &quot;main&quot; edu.stanford.nlp.util.ReflectionLoading$ReflectionLoadingException: edu.stanford.nlp.util.MetaClass$ClassCreationException: java.lang.ClassNotFoundException: edu.stanford.nlp.parser.lexparser.EnglishUnknownWordModelTrainer
    	at edu.stanford.nlp.util.ReflectionLoading.loadByReflection(ReflectionLoading.java:39)
    	at edu.stanford.nlp.parser.lexparser.BaseLexicon.initializeTraining(BaseLexicon.java:335)
    	at edu.stanford.nlp.parser.lexparser.LexicalizedParser.getParserFromTreebank(LexicalizedParser.java:800)
    	at edu.stanford.nlp.parser.lexparser.LexicalizedParser.trainFromTreebank(LexicalizedParser.java:226)
    	at edu.stanford.nlp.parser.lexparser.LexicalizedParser.trainFromTreebank(LexicalizedParser.java:237)
    	at ABravoDemo.main(ABravoDemo.java:35)
    Caused by: edu.stanford.nlp.util.MetaClass$ClassCreationException: java.lang.ClassNotFoundException: edu.stanford.nlp.parser.lexparser.EnglishUnknownWordModelTrainer
    	at edu.stanford.nlp.util.MetaClass.createFactory(MetaClass.java:353)
    	at edu.stanford.nlp.util.MetaClass.createInstance(MetaClass.java:370)
    	at edu.stanford.nlp.util.ReflectionLoading.loadByReflection(ReflectionLoading.java:37)
    	... 5 more
    Caused by: java.lang.ClassNotFoundException: edu.stanford.nlp.parser.lexparser.EnglishUnknownWordModelTrainer
    	at java.net.URLClassLoader$1.run(URLClassLoader.java:200)
    	at java.security.AccessController.doPrivileged(Native Method)
    	at java.net.URLClassLoader.findClass(URLClassLoader.java:188)
    	at java.lang.ClassLoader.loadClass(ClassLoader.java:303)
    	at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:301)
    	at java.lang.ClassLoader.loadClass(ClassLoader.java:248)
    	at java.lang.ClassLoader.loadClassInternal(ClassLoader.java:316)
    	at java.lang.Class.forName0(Native Method)
    	at java.lang.Class.forName(Class.java:169)
    	at edu.stanford.nlp.util.MetaClass$ClassFactory.construct(MetaClass.java:119)
    	at edu.stanford.nlp.util.MetaClass$ClassFactory.&lt;init&gt;(MetaClass.java:192)
    	at edu.stanford.nlp.util.MetaClass$ClassFactory.&lt;init&gt;(MetaClass.java:53)
    	at edu.stanford.nlp.util.MetaClass.createFactory(MetaClass.java:349)
    	... 7 more

How could I create a new model with this corpus ?


Content Type	Original Author	Original Content on Stackoverflow
Question	Surya	View Question on Stackoverflow
Solution 1 - Python	Franck Dernoncourt	View Answer on Stackoverflow
Solution 2 - Python	forivall	View Answer on Stackoverflow

Any tutorials for developing chatbots?

Python Problem Overview

Python Solutions

Solution 1 - Python

Solution 2 - Python

Proper REST formatted URL with date ranges

jQuery: How to detect window width on the fly?

Attributions