-The computational linguistics is now undoubtedly a well-developing and prospective field of study. As an intersection between linguistics as such and the computer science, it treats many problems of how to process the natural language as to make it applicable and easily transformative for the machines. The standard question we put ourselves when it comes to the interconnected areas of the natural language processing and the computer science, is whether we can teach a computer how to speak a natural language. The issue we will be thinking over in the paper, will be the way of treating the matter of standard computational linguistics problems we do encounter on a daily basis where it connects to the field of linguistics. In this paper, a novel approach to natural language processing by using the neural networks and object oriented approach is presented.
When we try to define what the interconnections between the natural language and the computer science really are we would have to take into account the fields of study both of the disciplines regard.
Linguistics, being a study of human language consists of almost seven thousand of languages there are in the world as the subjects of examination. It actually studies all the appearing linguistic aspects, that is to say: syntax, pragmatics and semantics. So a language, with all the aspects included, is being learned by us from a minimal input in our early childhood and serves us to communicate with each other more or less easily. A high variety of languages we do have nowadays causes also that there emerges some questions regarding the translation from one language into another. The problems to be sorted out in the field are pretty most frequently caused by the fact that the semantics of the language [applies also for semiotics, syntax and pragmatics] is being acquired by us, the users of the language, quite intuitively.
As the computers and machines do not function that way, we do meet problems like the one with how to automatically process the construction of the possesives pair of words, i.e. the photos of my friends that a user of a natural language would rather understand as pictures presenting the friends of the person who speaks than (as it could appear in the computer translation that the photos as the property of one’s firend’s).
Copyright c 2016 held by the author.
Natural language processing binds very strongly to the
subject of artificial intelligence. The idea of creating an
artificial consciousness expanded stream of science fiction
literature in the nineteenth century, and rapid technological
development aims to realize dreams of authors this type of
books. Already in the forties of the twentieth century, an
artificial neural network model has been designed. The one
which has numerous applications in life today. Especially,
neural networks [
Another group of important methods are heuristic
algorithms, which were created in order to find the maximum and
minimum values of optimized functions. Heuristics proved to
be a good option for finding solutions, not only for the problem
of searching the extremes of functions, but also in the graphics
processing [
Natural language processing is such an important subject
that can not only afford to develop the field of artificial
intelligence [
In these paper, I would like to present a novel approach to find the author of a longer text with the use of methods of artificial intelligence. The proposed model has been tested and described with regard to all its advantages and disadvantages.
MULTIDIMENSIONAL LOOK AT MODERN LINQUISTICS While studying the opportuinities of solving the problems computers do encounter when processing the languages we have to take into consideration also the relations between the words in the sentences. Even if a language is being treated intuitively, we ought to remember that the standard defining the quality of an enunciation is also the grammatical correctness. Like in the computer languages, in the natural ones as well, we have combinations that are either possible or not. Let’s take an example.
We have a complex phrase 1) 2) a) b) a) b)
Every time John goes out, he puts on a hat [he = John: possible]; Every time he goes out, John puts on a hat.
[he=John: possible].
The sentences above, therefore, show us very well that actually every native speaker of a language has naturally, with a quiet mininal imput from the childhood an intuition that tells him whether a phrase [a relation of words appearing in a sentence] could be correct and whether a grammatically and formally correct sentence could ever be constructed like this.
So what is interesting about the phrases we have been reflecting on is that there was no need of having taken syntax classes to know which of the sentences is apparently ungrammatical. Thanks to the knowledge we have nowadays, we are able to say that on the basis of the words of a natural language and the relations between them we can construct an infinite number of sentences that will be grammatically correct.
Even if we hear a sentence for the first time in our lives, we can suspect or just verify whether it is correct or not. The phrase: The six-headed CS84 Tbs grilled the blind octopus using a MAPA mug is surely correct when it comes to its grammar but not necessarily met by us ever before. That proves language functions intuitively.
III.
LINQUISTICS
We can check if the order and relation between the subject and predicate is as it should be. From among many problems of either the natural language processing and the computer science issues we would like to focus just on a particular part of them. To see how many chances are given us by the modern computer sciences and its processing the language we will be trying over the article to find an answer to the question of the classification problem. The problem we would like to sort out is: given two presidential speeches from the US election can I guess with high probability whose speech it is?
Naturally, when it comes to the classification problem there is no simply answer to the question like the one above. We can, however, assume for the purposes of our study that both presidential candidates speak about recurrent themes, also probably using reccurent words. Even if there is no exact answer to this, we will try make a few assumptions and create a model to try to accomplish what we suppose.
The assumption we will be making will be based on what we know for now about the presidential speeches. We will take into consideration two presidential speeches of Barack Obama and of Mitt Romney.
”And we salute the people of Paris for insisting this crucial conference go on; an act of defiance that proves nothing will deter us from building the future we want for our children (. . . ) I want to show her passionate, idealistic young generation that we care about their future. (. . . ) This summer, I saw the effects of climate change firsthand in our northernmost state, Alaska, where the sea is already swallowing villages and eroding shorelines; where permafrost thaws and the tundra burns, where glaciers are melting at a pace unprecedenten in modern times. And it was a preview of one possible future.” [Barack Obama, First Session of COP 21, Nov 30, 2015].
As we have seen in the Obama’s speech the word future appears three times. Let’s now have a look at Mitt Romney’s speech.
”(. . . ) They came not just in pursuit of the riches of this world but for the richness of this life. Freedom. Freedom of religion. Freedom to speak their mind.. Freedom to build a life. And yes, freedom to build a business. With their own hands.” [Mitt Romney, Republican Convention, Aug 08, 2012] The assumption we can make for now is that American people have been hearing Romney’s speaking very frequently about freedom in various contexts, like: freedom of religion, freedom to speak your mind, freedom to build a business, freedom to build a life.
Followingly, what is specific for Obama’s speeches is the notion of future. He speaks about the future of America, IV. PREPROCESSING OF THE ELEMENTS OF DISCOURSE
To prepare long enunciations for the purposes of an analysis of the problems of the AI, we have to represent them in the simpliest possible form, applying there some number values. To show how that actually functions we will be serving ourselves with the Bayes theorem which defines directly how the conditional probability works and can be seen as a way of understanding how the probability that a theory is true is influenced by an evidence that appears there for the first time. Let’s illustrate it with a theorem argmaxP r(cjw) = = argmax
P r(w) = argmaxP r(wjc)P r(c);
P r(wjc)P r(c) = where P r(cjw) stands for the probability of appearing of a particular word w in a particular class c of objects.
The Bayes theorem ([
In order to increase the precision of ANN, many learning
algorithms of this type of network are created. One of these
algorithms is the backpropagation algorithm, which works on
minimizing the error function and modify weights from the
output layer to the first hidden layer [
i(1 : i)(!i i) X j2
i) for output layer wji i for hidden layer where
is the output value and ! is the expected value. (3) (4) (5) (6) i is ; (7) G(x) =
exp 2 2 )2 ; is the height of the function, and is the shift on where y-axis.
The Gaussian blur permits us approximate the values as to assess the probability of an expression appearing in the demanded conditions. The calculated values allow us create a vector representing the enunciation desired. That could be showed by such a theorem
[G(y0); G(y1); : : : ; G(yn 1); id]; where G(yi) is stand for values calculated, in accordance with the equation (3), and id means the author’s numerical identifier.
V.
NEURAL NETWORK
Already in the forties of the twentieth century, the author
of [
ANN is composed of three types of layers – the input, hidden and output. The input layer is responsible for the acceptance of teaching vector, and output for return a result of the network. Hidden layers are located between the input and output, they are responsible for creating a deeper network in order to obtain better results.
Each layer is constructed of neurons, wherein each neuron
of one layer is connected to each neuron in the next layer.
Neuron is the smallest object of neural network. The data
enters the neuron through mergers which have a certain weight.
In the neuron, all input values are rescaled by the activation
function. The value after scaling is sent to other neurons along
outbound connections to the next layer. As activation function
selected a bipolar sigmoid function [
2 1 + exp( x) ; where is a parameter in (0; 1]. Activation function graph is shown in Fig. 6.
Fig. 5: Knowledge representation using Sammon mapping. VI. PROPOSED MODEL OF AUTHOR’S IDENTIFICATION
The problem of a person verification on the basis of a longer text requires not only the extraction of the characteristics of his speech, but also correct classification. For this purpose, the proposed model consists several stages. At the beginning, the statement is entered into a computer, where it statement is processed according to Sec. IV. The next step of action has two paths. In the first one, the sample is stored in the database. When the database contains a sufficient number of samples, the neural network is trained using the knowledge contained in the database. The second path of action is to classify the samples by the neural network.
A model of such a system is shown in Fig. 3.
EXPERIMENTS
The proposed solution has been tested by the use of 200 samples – 100 samples per person. Each sample contained a fragment of statements about the future, taking into account up to 60 words. For the purposes of minimizing the time of neural networks learning, each sample contained 15 components. A neural network was composed of 4 layers input layer composed of 15 neurons;
output layer consisting of one neuron.
To train the network, the samples were divided into two groups (training and verifying – 80% : 20%). The problem of classification has been shown in Fig. 5 using Sammon mapping – based on this interpretation of the spread of knowledge it can be seen that there can not be any easy way to separate the samples into two groups, so the problem of classification is extremely difficult. The network was trained to obtain the error of the 0:24 - error learning graph is shown in Fig. 4. In order to verify the operation of the classifier, each sample was given to the input of the network. In consequence of the operation, the system indicate the author correctly for 103 samples, which results indicates in a efficiency at about 72%.
CONCLUSION
The subject of the research was to prove the possibilities of processing the natural language by the methods of computational linguistics. We have shown that, given two different longer texts, we are able to identify their authors exclusively on the basis of the words used - with a minimal risk rate possible (error - 0:24). After entering the data into the computer, the use of the database’s contents needed to be classified by the neural network. Having done so, we processed two independent author’s speeches Barack Obama’s and Mitt Romney’s addresses and therefore, helped by the neural networks, we could evaluate whose a text is just on the basis of the samples given and after comparing the data we had with the one contained in the database. The outcomes then of such an experiment are of significant help when it comes to the examination of idiolects and distinctive, personal styles of constructing a discourse. It – followingly – contributes to the development of the Artificial Intelligence as it is the computer to identify the authorship of a text. [5] R. Johnson and T. Zhang, “Effective use of word order for text categorization with convolutional neural networks,” arXiv preprint arXiv:1412.1058, 2014.