5
The website clearly does not fulfill the requirements of Reproducible Research. The algorithm is not Open Source and the implementation isn't available for inspection.
There are more problems:
The Website can go away without prior notice
The implementation of the algorithm may change without notice
When you cannot find an Open Source implementation of the ...
answered May 23 '18 at 12:36
jk - Reinstate Monica
22.1k3 gold badges51 silver badges94 bronze badges
5
The linked page is prominently open source. You can probably download the sources to your own computer and tweak the parameters to make it speak more slowly. It links to http://www.masswerk.at/mespeak/ which displays detailed controls for amplitude, pitch, speed, word gap, etc. The same controls are hardcoded into mespeak.js in the fragment $speed=175;$amp=...
4
Sentiment Classification is a still developing field so I don't feel comfortable saying that there is a right or a wrong way to approach your problem. Your approach would theoretically work but it is a bit simplistic. Consider the following:
First, how will you choose your word list? Will it be manually compiled? How will you make sure it is representative ...
4
Although doing so it not with caveats, it is not hard to covert IPA to speech today. As other answers suggest, the free software/open source program eSpeak has been able to synthesis speech from something like IPA for years but its notation for describing phonemes is different. As a result, one must first convert from IPA to eSpeak's notation. Once that's ...
4
This Theano tutorial provide a dialogue state tracking systems (slot filling) with a graphical model based (recurrent neural networks) dialogue state representation.
You probably need a decently large training set though: on the Fourth Dialog State Tracking Challenge (DSTC4) last year, we (and other teams) unsuccessfully tried some neural networks but in ...
4
"The hardest" is difficult to say. I will say that online translators often seem to have especial trouble with Japanese, as its often implicit anaphors (subjectless sentences and so forth) are extremely difficult for a machine algorithm to infer, so you often get lots of confused pronouns, mixed up genders, etc.
4
It depends on generally syntactic structure of two languages and -yes- some pragmatic issues. To answer your question,
the easiest languages to be translated into English by any automatic translation software are the ones which are nearly close to syntactic structures of English such as German, Chinese blah blah.
the hardest ones are the ones which are far ...
3
If I'm remembering correctly, according to Atkins and Rundell (2008) [1], many projects build their own software, but there are also commercial packages such as TshwaneLex (http://tshwanedje.com/tshwanelex/).
Atkins and Rundell (2008) go into some detail about the requirements of such software, so you may find their book helpful. They mention that there are ...
3
The probably most widely used concordance software is Laurence Anthony's AntConc. It has a range of functions beyond producing concordances, such as determining word counts and collocations. It can also handle regular expressions. AntConc is free and runs on a number of different platforms. I regularly recommend it to students, use it in teaching, and know ...
3
The example you give displays three grammatical aspects to deal with. First, the subject should agree with the verb by person and number, which is relevant in present tense: “John eats” and not “John eat”. Second, you need to construct the verb phrase correctly according to the required tense/aspect/voice, i.e., present-perfect-progressive will be “John has ...
2
This book by Pang and Lee is a classic reference on the topic. To get a good overview of the topic, look at the book's website which has links to presentation slides.
I see that you want to implement the software on your own, and you want to (re-)invent algorithms too. The reasons behind these is not entirely clear. If you are writing a research paper or a ...
2
A single program that supports both alignment and adding morphological annotation is impossible in the next 10 years.
Having a fully annotated and sentence-aligned version is everybody's goal. I'm working on a Sanskrit-Russian corpus and so I build upon http://kjc-fs-cluster.kjc.uni-heidelberg.de/dcs/index.php?contents=texte and add Russian parallel texts.
...
2
yourpronunciationlab can do that. This app automatically replays the sample, records your repetition, replays the sample and the repetition together then the process starts from the beginning.
2
You can try the Spanish stemming algorithm found at http://snowball.tartarus.org/algorithms/spanish/stemmer.html.
2
Take a look at below:
Deep Linguistic Analysis for Topic-level Analysis
Bitext’s API uses Deep Linguistic Analysis based on grammars, which allows for opinion analysis not only at the sentence level, but also at the phrase level within the sentence. This is possible because the syntactic analysis identifies the different phrases (noun phrases, adjective ...
2
MarkDown is not very standardized, so you can always add "yet another MarkDown dialect". Perhaps the easiest way to support a language such as you describe, is to write several regular expression rules to transform it straight to HTML+CSS. There are many MarkDown-to-HTML programs you could start from, too.
One difficulty that arises in trying to design such ...
2
espeak does a lot of what you are after, it seems to deal with both arbitrary words as well as having a list of exceptions. Though how complete this list of exceptions is remains to be seen.
espeak -q -v de --ipa warm
It seems to be the case that german orthography is pretty regular... though I don't fluently speak german so don't trust me.
Worth noting: ...
2
If you are familiar with Python an alternative for Flex would be NLTK's tokenizer:
>>> from nltk.tokenize import TweetTokenizer
>>> tknzr = TweetTokenizer()
>>> s0 = "This is a cooool #dummysmiley: :-) :-P <3 and some arrows < > -> <--"
>>> tknzr.tokenize(s0)
['This', 'is', 'a', 'cooool', '#dummysmiley', ':'...
2
So in researching the question, I ended up finding an answer.
While it does not give me full control over the pausing, adding a period after each block of phonetic characters added the perfect amount of spacing, thus:
<phoneme alphabet="IPA" ph=" ˈaɪ. ˈiː. dˈiː.z ">IEDS</phoneme>
Works perfectly.
1
I assume that you are not asking how you, yourself, can talk slowly, you want to hear recordings stretched out. The first thing is to get the sound file itself, which in the case of Forvo is not trivial, but not impossible. Let's just assume that you have a sound file on your computer. The program Speech Analyzer has 'slowing down' as a built-in slider-type ...
1
AntConc won't ever be comparable to professional tools like Python or R. And it's not possible to know what the software is doing, or to adapt it, since it's not open source (it's freeware, I believe).
You could also create some GUIs on the top of Python packages like nltk and have the functions that you use at hand.
1
A software programme you can also use is WordSmith Tools: http://www.lexically.net/wordsmith/version6/index.html
This website has a list of software that can be useful for linguistics:
http://linguistlist.org/sp/GetWRListings.cfm?WRAbbrev=Software
There are multiple programmes listed for corpora exploration, which probably also includes freeware.
1
The parser you linked to has been rewritten in Java and improved. The latest software can be downloaded from Stephen Clark's C&C tools page. The old C++ version can be downloaded from there too, which comes with Boxer.
As for state-of-the-art tools, I can't recommend anything, but SIGPARSE and SIGSEM are good starting points to find them.
1
These are some examples I've found myself:
https://github.com/jeremyfix/dstc (rule-based, advanced visualization)
https://github.com/CallumMain/DNN-DST (Deep Neural Net-based)
https://github.com/UFAL-DSG/xtrack2 (Recurrent Neural Net-based)
https://github.com/UFAL-DSG/alex/blob/master/alex/components/dm/dstc_tracker.py (Bayesian Discriminative tracking)
...
1
You do not have to do much if you are going to make an Extractive summarization.
Do stop-word Removal
Do stemming (optional)
Count all words occurrences
Extract sentence(s) that contains the word with most frequency in the text
This is the basic of text summarization which suits for emails very well, I guess.
1
Here's a tool I happened upon recently: Context-controllable Content Summarization
And this looks to be related background information:
Generating content summarizations similar to the way a human might do so
Word2Vec is based on an approach from Lawrence Berkeley National Lab
https://code.google.com/p/word2vec/
1
There are two pieces of software that I know of that can show Spectograms.
Praat is one, and is the more powerful of the two, but may have a steeper learning curve. I think that they are aimed more at linguists who analyse a language rather than those trying to learn a language.
The other is Speech Analyser Available for windows PC's though there may be a ...
1
I'm working on American English pronunciation software Saundz that is intended to help ESL learners improve their pronunciation. It features a 3D animated teacher who demonstrates movements of vocal organs, so that students can see tongue placement, jaw movements, as well as air stream (with nasals, for example). There are 40 basic American English sounds ...
1
A bit late as a response - I've just stumbled on this. But I believe SPAD-T was released as the open-source DTM-VIC available here http://www.dtmvic.com/05_SoftwareE.html (10 April 2016)
1
The most challenging (not to say, practically impossible with current development of NLP) is to translate between languages that are built around completely different concepts.
As for Indo-European languages or languages like Chinese, Georgian, Swahili, the concepts are similar. You have nouns described by verbs and adjectives, those are described by ...
Only top voted, non community-wiki answers of a minimum length are eligible
