3.1 Proposed Approach
In the following sections, the approach used in the paper is described and explained in detail. As shown in Fig. 1 we start with the section of data collection, then explain the data pre-processing steps. The next section talks about document vector creation and the advantages and disadvantages of the approaches.
3.2 Data Preparation
This section describes in detail the proposed approach to tackle the purchase intention detection issue. The data collection and annotation processes are described before explaining the data pre-processing and transformation as per the requirements of analytical models.
3.2.1 Data Collection and Annotation
Due to the unavailability of annotated Twitter corpora regarding the detection of purchase intention, manually annotated data set was created for this study. The product picked for our task was the iPhone X. The tweets were downloaded using a web crawler developed by JohnBakerFish [14] which crawled the website to collect the data [15]. Subsequently 3,200 randomly picked tweets were manually annotated using the basic criterion defined in Table 1. Each tweet was assigned a label from two people. In case of differing judgement, a third person reviewed the tweet before assigning a final label.
Table 1: Criteria for Labelling of tweets
![](data:image/png;base64,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)
Purchase intention (PI) can be defined as a situation where a consumer is willing and intends to make online transactions [16]. For the purpose of this research paper a purchase intention shown by a user in a Twitter tweet is represented by the tweet having action words like buy, want, and desire associated with the product the user intends to buy. After the annotation process, we had 661 tweets with Purchase Intention and 457 tweets with No Purchase Intention. The remaining tweets were too ambiguous or neutral.
3.2.2 Data Pre-processing
Figure 2 shows how the tweets were processed using the following techniques in sequential order. At first, the text of the tweets was converted into lower case. Word correction was applied to all tweets to remove spelling errors. URLs were also removed from all tweets as they would not be contributing towards the model’s learning. Contractions like can’t and you’re were converted to their full form. The text was then passed through punctuations and special characters filters, as the text may contain unwanted special characters, spaces, tabs, and etcetera which have no significance in text classification. In some cases, the tweets had negation modifiers, like the word “not” which modifies the meaning of the associated word. Handling Negation Modifiers is important as they may modify the sentiment conveyed by the associated word [17]. The negation modifiers were handled by appending a custom defined keyword “NOT_” to each word appearing between a negation and following punctuation [18]. Negation modifiers were only used for bag-of-word approaches.
Next step was stop-words removal since the tweets also contain frequent words which are part of the regular sentence and grammar but do not contribute to the meaning of the sentence. The removal of stop words like “the”, “a” and “an” also reduces the dimensions of the vocabulary. The words “phone” and “x” had the most occurrence and were independent of purchase intention, so they were removed from the corpora. Finally, the tweets were subjected to stemming. Stemming reduces a word to a root form by removing inflection through dropping unnecessary characters, usually a suffix. There are several stemming models such as Porter and Snowball [19]. For this study, Porter's Stemmer is used which is available with NLTK, a library of Python [20]. Some of the experiments were conducted with lemmatization. Lemmatization, unlike Stemming, also ensures that the root word belongs to the language. In Lemmatization, the root word is called lemma. A lemma is the canonical, dictionary, or citation form of a set of words [21]. Lastly, the tweets were broken down into word tokens using NLTK’s tokenizer which utilizers the TreeBank tokenizer algorithm.
3.3 Formation of Document Vectors
In order to pass the tweets to the model, they had to be converted to a numerical representation. A classical technique used was BOW (Bag of Words) representation. Bag of Words takes all unique words from the corpus to form a vocabulary. Words are assigned a sequence number and used as the index of the vector. In binary BOW, a one-hot encoding approach is taken and if the word occurs in the document, a 1 is marked. Otherwise, 0 stays. We used Count BOW instead of binary BOW which just marks the words’ index position with its number of occurrences in a document, instead of a 1. TF-IDF (Term Frequency – Inverse Document Frequency), the second BOW representation used, works similarly. Instead of assigning a 1 or count at the index position of the word, a ratio is assigned. Term Frequency is the number of occurrences of a word in a document, divided by the total number of words in said tweet. Inverse Document Frequency is the number of documents where a word occurs, divided by total number of documents. This ratio has an additional benefit of assigning smaller values to words that are common among all documents. These words usually do not provide any distinction capability [22].
The drawback of using classic techniques mentioned above is that they do not capture any contextual information in their word embeddings. No information regarding a word’s position is used and the words are clumped together in no specific order, hence the name Bag of Words. Using n-grams some-what mitigates this problem, but it has tremendous computational and practical limitations. Another approach to generate word representations is to use a co-occurrence matrix of word pairs (or windows) in the corpus [23]. Dimensionality reduction (Principal Component Analysis, Singular Value Decomposition) can be used to reduce the size of the matrix. While this approach can work, we decided to opt for a more complex approach (GloVe) that enhances this method.
A more modern approach is using distributed representation models like Word2Vec and GloVe to generate word embeddings in high dimensions. Both models work by taking a sequence vector (one-hot encoding of document) as input. The input vector undergoes multiplication with an embedding matrix of the form m x n where m is the vocabulary size and n are the number of dimensions specified to encode the word, a parameter that we can set. The output is a vector for each input word in the document. This is a tremendous improvement from the heavily sparse matrix that would have been created had we used a bag of words approach. Additionally, these embeddings maintain contextual information that should help the model capture more complex patterns and perform better.
Word2Vec embeddings come from a Neural Net that is trained to predict a word from a context window (Continuous Bag of Words) or predict context from a word (skip-gram).
GloVe embeddings come from a log bilinear regression model that is trained to generate embeddings such that they are associated with the ratio of co-occurrence matrix for word and context.
Since these models are trained to predict words from context (skip-gram predicts context words), the resulting vector representations hold semantic as well as syntactic information [24].
We used both Word2Vec and GloVe based word embeddings to generate word vector representations. Since these techniques generate vectors on a word level and not a document level, we were left with few options. The first option was to perform an average of all word vectors within a tweet and use that as an input feature to our Machine Learning Model. This was not promising. The second option was to flatten all word vectors in each tweet to become one long vector. This would lead to a curse of dimensionality and degraded performance of models. Positional information would be lost as well. The last option was to deploy a deep learning model of Recurrent (RNN) Architecture. These models are designed to take input sequences concurrently and are heavily utilised for NLP tasks. This allowed us to feed the embeddings of each tweet directly to the model for classification. We used the LSTM (Long Short Term Memory) network instead of the standard RNN as it overcomes the problems of vanishing and exploding gradients [13]. LSTM are also more capable of learning longer term dependencies.
The inputs for LSTM were generated sequence tokens of a fixed size of 55. Right padding was done so that all tweets had the same size.
3.4 Modelling
The problem of identifying purchase intent is one of classification. Therefore, we utilized supervised classification models to predict intent. Five most frequently used models namely, Support Vector Machine, Naïve Bayes, Logistic Regression, Decision Tree and Artificial Neural Network, were trained and tested using the cleaned data [25]. LSTM network was used for creating models from word embeddings of Word2Vec and GloVe.
3.5 Machine Learning Approaches
The first model used was the multinomial Naive Bayes classifier. The classifier was configured with Laplace smoothing to prevent zero probabilities for features missing from the training samples. The Prior probabilities of the features were used rather than using a uniform prior probability.
The second model used was the Support Vector Machine classifier. A Linear SVM was implemented with the penalty of an error set to 1 and trained using cross-validation.
The third model was the Logistic Regression having the inverse of regularization strength coefficient set to 1 for stronger regularization and the maximum number of iterations to convergence set to 100. For optimization of the model the liblinear algorithm was used, as it is best suited for small datasets.
The Decision Tree classifier was also tested. Gini Index was used to determine the quality of each split. The highest accuracy was obtained by splitting the internal node using at least seven samples.
The Multi-Layer Perceptron model was the last ML model, and it was optimized using stochastic gradient descent (SGD) with Cross Entropy loss. Rectified linear unit (ReLU) was used as the activation function for the hidden layer, as it is faster to compute [26]. It also converges faster and is not prone to vanishing gradients unlike sigmoid and tanh activations. The learning rate schedule for weight updates was kept constant. Four hidden layers with the first hidden layer having 50 neurons, the second one having 20 neurons, the third one having 10 neurons and the last hidden layer having 5 neurons. The output layer had only one node for binary classification. Training was done for 50 epochs.
3.6 Deep Learning Approaches
For the Word2Vec model, an embedding layer was used with 300 embedding dimensions. For the GloVe model, an embedding layer with 200 dimensions was used [27]. Both embedding layers are from the standard pre-trained models that were trained on corpus. Pre-trained models are especially helpful as they are supposed to contain more accurate representations of words and sentences.
An LSTM layer was added on top with 32 units and a dropout of 0.2 to prevent overfitting. Sigmoid activation was used for the output layer as the model was binary classification. Model was trained using a cross entropy loss function and the optimizer used was stochastic gradient descent.