You are reading the article Build Your Own Fake News Classifier With Nlp updated in December 2023 on the website Kientrucdochoi.com. We hope that the information we have shared is helpful to you. If you find the content interesting and meaningful, please share it with your friends and continue to follow and support us for the latest updates. Suggested January 2024 Build Your Own Fake News Classifier With Nlp
This article was published as a part of the Data Science BlogathonThe major objective of watching or reading news was to be informed about whatever is happening around us. There are several social media platforms in the current modern era, like Facebook, Twitter, Reddit, and so forth where millions of users would rely upon for knowing day-to-day happenings. Then came the fake news which spread across people as fast as the real news could. Fake news is a piece of incorporated or falsified information often aimed at misleading people to a wrong path or damage a person or an entity’s reputation.
Characteristics of Fake News:
Their sources are not genuine.
May or may not have grammatical errors.
Seems too good to be true.
Mimics the real headlines and twists the story.
As humans, when we read an article, we could somehow understand its context by interpreting its words. Given today’s volume of news, it is possible to teach computers how to read and understand the difference between real and fake news using NLP techniques. All you need here are the appropriate Machine Learning algorithms and a dataset.
Let’s have a quick look at the workflow for the Fake news classifier model!
1.Data Collection:
The process of gathering information from various and all possible resources regarding a particular research problem. This information is stored in a file as the dataset and is subject to various techniques like testing, evaluation, etc.
2.Data Cleaning:Identification and removal of errors if any in the gathered information. This process is carried out mainly to improve the dataset’s quality, make it reliable, and provide accurate decision-making processes.
3.Data Exploration/Analysis:Various visualization techniques are carried out here to understand the dataset in terms of its characteristics namely, size, quantity, etc. This process is essential to better understand the nature of the dataset and get insights faster.
4.Data Modelling:The process of training the dataset using one or more ML algorithms to tune it according to the business need, predict or validate it accordingly.
5.Data Validation:The method of tuning the hyperparameters before testing the model. This provides an unbiased evaluation of a model fit done on the training dataset.
6.Deployment:Integrating an ML model into an existing environment to make more practical business decisions based on the dataset.
Python ImplementationThe dataset downloaded from Kaggle has the following attributes: Id, Title, Author, Text, and the Label (where 1 is unreliable and 0 is reliable).
import pandas as pd df = pd.read_csv('news.csv') #df.head()Having downloaded and read the dataset, we now have to check and classify its features (or attributes), this step generally depends on what the dataset contains and what we are about to analyze. Thus, in this classifier model, we’ll be taking the “Id”, “Title” and “Author” columns as Independent features and the “Label” column as a dependent feature.
#get the independent features X = df.drop('label',axis =1) #X.head() #get dependent variables y = df['label'] #y.head()In total, there are about 20800 records.
df.shapeOutput:
The dataset is checked for any missing values and those are dropped.
from sklearn.feature_extraction.text import CountVectorizer,TfidfVectorizer,HashingVectorizer df = df.dropna() #df.head(10)Now we’ll create a copy of this dataset and also reset its index values.
messages = df.copy() messages.reset_index(inplace = True) #messages.head(10)This part of the code is for Stopword removal and Stemming. Here, the corpus is an array in which we have appended all the titles of the news.
import re from nltk.corpus import stopwords from nltk.stem.porter import PorterStemmer ps = PorterStemmer() corpus = [] for i in range(0, len(messages)): review = re.sub('[^a-zA-Z]', ' ', messages['title'][i]) review = review.lower() review = review.split() review = [ps.stem(word) for word in review if not word in stopwords.words('english')] review = ' '.join(review) corpus.append(review) #corpus[3]Applying count vectorizer (also known as the “Bag of Words”). Maximum features passed here are 5000.
#Applying countvectorizer #Creating the bag of words model from sklearn.feature_extraction.text import CountVectorizer cv = CountVectorizer(max_features = 5000,ngram_range=(1,3)) X = cv.fit_transform(corpus).toarray() X.shape #(18285, 5000) y =messages['label']Dataset is now split into train and test. The first 20 features are displayed here.
#Divide the dataset into train and test from sklearn.model_selection import train_test_split X_train,X_test,y_train,y_test = train_test_split(X,y,test_size = 0.33,random_state = 0) cv.get_feature_names()[:20]Output:
‘access pipelin protest’]
We’ll now get the parameters as the next step.
cv.get_params()Output:
‘vocabulary’: None}
The array is converted into dataframe.
count_df = pd.DataFrame(X_train,columns = cv.get_feature_names()) #count_df.head()Using the matplotlib function, lets plot a confusion matrix. This part of the code and the plot_confusion_matrix function will be used in all the upcoming algorithms for its training and accuracy prediction purpose.
import matplotlib.pyplot as plt def plot_confusion_matrix(cm, classes, normalize=False, title='Confusion matrix', cmap=plt.cm.Blues): plt.imshow(cm, interpolation='nearest', cmap=cmap) plt.title(title) plt.colorbar() tick_marks = np.arange(len(classes)) plt.xticks(tick_marks, classes, rotation=45) plt.yticks(tick_marks, classes) if normalize: cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis] print("Normalized confusion matrix") else: print('Confusion matrix, without normalization') thresh = cm.max() / 2. for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])): plt.text(j, i, cm[i, j], horizontalalignment="center", plt.tight_layout() plt.ylabel('True label') plt.xlabel('Predicted label') Multinomial Naive Bayes AlgorithmThis algorithm considers the integer feature counts where it represents how often or how many times a specific word appears. It is much suitable for the classification of discrete features. i.e. word counts for classifying a text. The Prediction Accuracy we’ve got here is 90%.
from sklearn.naive_bayes import MultinomialNB classifier = MultinomialNB() from sklearn import metrics import numpy as np import itertools classifier.fit(X_train,y_train) pred = classifier.predict(X_test) score = metrics.accuracy_score(y_test,pred) print("Accuracy: %0.3f"%score) cm = metrics.confusion_matrix(y_test,pred) plot_confusion_matrix(cm, classes=['FAKE', 'REAL'])Output:
Accuracy: 0.902
Confusion matrix, without normalization
classifier.fit(X_train, y_train) pred = classifier.predict(X_test) score = metrics.accuracy_score(y_test, pred) scoreOutput:
y_train.shape #(12250,)pred #
array([1, 1, 1, ..., 0, 0, 1])
Passive Aggressive Classifier AlgorithmThe PAC algorithm responds aggressively to incorrect predictions and remains passive for the correct predictions. By using this algorithm, the accuracy is 92%.
from sklearn.linear_model import PassiveAggressiveClassifier linear_clf = PassiveAggressiveClassifier(max_iter=50) linear_clf.fit(X_train,y_train) pred = linear_clf.predict(X_test) score = metrics.accuracy_score(y_test,pred) print("Accuracy: %0.3f"%score) cm = metrics.confusion_matrix(y_test,pred) plot_confusion_matrix(cm,classes = ['FAKE Data','REAL Data'])Output:
Accuracy: 0.920
Confusion matrix, without normalization
Multinomial Classifier with Hyperparameter (alpha)It’s none other than the Multinomial Naive Bayes algorithm we trained earlier but here we’ll be adding a hyperparameter for it and for each alpha value, the corresponding scores are calculated. At alpha = 0.3, the score seems to be maximum.
The more negative the coefficient value, the word is used in fake news!
classifier=MultinomialNB(alpha=0.1) previous_score=0 for alpha in np.arange(0,1,0.1): sub_classifier=MultinomialNB(alpha=alpha) sub_classifier.fit(X_train,y_train) y_pred=sub_classifier.predict(X_test) score = metrics.accuracy_score(y_test, y_pred) classifier=sub_classifier print("Alpha: {}, Score : {}".format(alpha,score))Output:
Alpha: 0.0, Score : 0.8903065451532726
Alpha: 0.1, Score : 0.9020712510356255
Alpha: 0.2, Score : 0.9025683512841757
Alpha: 0.30000000000000004, Score : 0.9024026512013256
Alpha: 0.4, Score : 0.9017398508699255
Alpha: 0.5, Score : 0.9015741507870754
Alpha: 0.6000000000000001, Score : 0.9022369511184756
Alpha: 0.7000000000000001, Score : 0.9025683512841757
Alpha: 0.8, Score : 0.9015741507870754
Alpha: 0.9, Score : 0.9017398508699255
#get features names feature_names = cv.get_feature_names() classifier.coef_[0]array([ -9.10038883, -8.62276128, -9.10038883, …,
–10.79498456, -8.91467169, -9.32864749])
And yay! Here you can see we have classified the most real and most fake news based on their coefficients. The fake news classifier model we just implemented has worked out pretty well 😀
#Most real sorted(zip(classifier.coef_[0], feature_names), reverse=True)[:20]Output:
(-5.862110622807369, ‘america’)]
#Most fake sorted(zip(classifier.coef_[0], feature_names))[:20]Output:
(-10.794984555596727, ‘american new’)]
If you’re an enthusiast who is looking forward to unravel the world of Generative AI. Then, please register for our upcoming event, DataHack Summit 2023.
ResourcesThe media shown in this article are not owned by Analytics Vidhya and are used at the Author’s discretion.
Related
You're reading Build Your Own Fake News Classifier With Nlp
This Could Make Creating Convincing Fake News Videos Too Easy
This could make creating convincing fake news videos too easy
Fake news is a big and thorny subjects these days and there is no indication that it’s going away any time soon. In fact, it might get worse before it gets better. Countless people are already easily duped by legit-looking text and professionally photoshopped images. But what if they can be deceived by video as well? While the machine learning algorithms developed by University of Washington researchers have more lofty goals in mind, this “lip-syncing tech” could also be used to prank and mislead people in the future.
Imagine taking footage from one video, say an interview of former US president Barack Obama, and mixing it with audio from a completely different video, like a totally different interview. While technically possible, the results are less than convincing. But using this new machine learning technique, University of Washington researchers have been able to produce such a mix that eerily looks like the real deal.
The problem with mixing video with different audio is in the way the mouth moves. Humans have an uncanny talent for detecting such out of sync or even fake mouth movements. Unsurprisingly, that is part of the aversion better known as “uncanny valley”. This new research fixes all that and in a more efficient way.
Audio to video conversion, as it is called, has been around for quite a while now, but almost all techniques involved multiple people being filmed to speak the same sentences repeatedly while cameras and computers capture the changes in the shape of the mouth and associate those with certain sounds. Not exactly ideal or cheap. The UW researchers, instead, utilized machine learning and fed existing videos to a neural network, which then proceeded to do the same analysis of mouth shapes and their associated sounds. You no longer need to have live actors on hand but can simply use tons of recorded videos to achieve the same effect.
That’s only half of the process though. The second half involves analyzing the “fake” audio input and then create mouth shapes corresponding to the sounds in the audio and superimpose them on the target video. The results are both impressive and worrying and many people probably won’t be able to tell that the resulting video is anything but legit.
The researchers envision that this technology could be used for things like video conferencing, where video can “catch up” with audio despite the lag, or for education, where historical figures can be made to look like delivering speeches recorded only in audio. But it can also be used to create fake videos of prominent figures saying something they might have said long ago or not at all.
The good news is that the machine learning agent is currently fixated on Mr. Obama. That is, it is the only thing it knows about, thanks to the over abundance of data, like video interviews, available. The second is that the researchers might also have a way to “reverse” the process to tell whether a video is fake or not.
SOURCE: University of Washington
Best Books To Expand Your Nlp Knowledge
The abundance of knowledge and resources can be at times overwhelming specifically when you are talking about new age technologies like Natural Language Processing or what we popularly call it as NLP. When trying to educate yourself, you should always choose resources with solid base and fresh books to impart unprecedented package of learnings. Here is the list of
Speech and Language ProcessingAuthors: Daniel Jurafsky and James H. Martin This is the second edition and Jurafsky and Martin are working on the third with a targeted completion later this year. View a draft on Jurafsky’s Stanford web page.
Natural Language UnderstandingAuthor: James Allen This book is another introductory guide to Natural Language Processing and considered a classic. While it was published in 1994, it’s highly relevant to today’s discussions and analytics activities and lauded by generations of NLP researchers and educators. It introduces major techniques and concepts required to build NLP systems, and goes into the background and theory of each without overwhelming readers in technical jargon.
Handbook of Natural Language ProcessingAuthors: Nitin Indurkhya and Fred J. Damerau This comprehensive, modern “Handbook of Natural Language Processing” offers tools and techniques for developing and implementing practical NLP in computer systems. There are three sections to the book: classical techniques (including symbolic and empirical approaches), statistical approaches in NLP, and multiple applications—from information visualization to ontology construction and biomedical text mining. The second edition has a multilingual scope, accommodating European and Asian languages besides English, plus there’s greater emphasis on statistical approaches. Furthermore, it features a new applications section discussing emerging areas such as sentiment analysis. It’s a great start to learn how to apply NLP to computer systems.
The Handbook of Computational Linguistics and Natural Language ProcessingAuthors: Alexander Clark, Chris Fox, and Shalom Lappin Similar to the “Handbook of Natural Language Processing,” this book includes an overview of concepts, methodologies, and applications in NLP and Computational Linguistics, presented in an accessible, easy-to-understand way. It features an introduction to major theoretical issues and the central engineering applications that NLP work has produced to drive the discipline forward. Theories and applications work hand in hand to show the relationship in language research as noted by top NLP researchers. It’s a great resource for NLP students and engineers developing NLP applications in labs at software companies.
The Oxford Handbook of Computational LinguisticsAuthor: Ruslan Mitkov This handbook describes major concepts, methods, and applications in computational linguistics in a way that undergraduates and non-specialists can comprehend. As described on Amazon, it’s a state-of-the-art reference to one of the most active and productive fields in linguistics. A wide range of linguists and researchers in fields such as informatics, artificial intelligence, language engineering, and cognitive science will find it interesting and practical. It begins with linguistic fundamentals, followed by an overview of current tasks, techniques, and tools in Natural Language Processing that target more experienced computational language researchers. Whether you’re a non-specialist or post-doctoral worker, this book will be useful.
Foundations of Statistical Natural Language ProcessingAuthors: Christopher Manning and Hinrich Schuetze Another book that hails from Stanford educators, this one is written by Jurafsky’s colleague, Christopher Manning. They’ve taught the popular Natural Language Processing introductory course at Stanford. Manning’s co-author is a professor of Computational Linguistics at the German Ludwig-Maximilians-Universität. The book provides an introduction to statistical methods for NLP and a decent foundation to comprehend new NLP methods and support the creation of NLP tools. Mathematical and linguistic foundations, plus statistical methods, are equally represented in a way that supports readers in creating language processing applications.
Natural Language Processing with Python: Analyzing Text with the Natural Language ToolkitAuthors: Steven Bird, Ewan Klein, and Edward Loper This book is a helpful introduction to the NLP field with a focus on programming. If you want have a practical source on your shelf or desk, whether you’re a NLP beginner, computational linguist or AI developer, it contains hundreds of fully-worked examples and graded exercises that bring NLP to life. It can be used for individual study, as a course textbook when studying NLP or computational linguistics, or in complement with artificial intelligence, text mining, or corpus linguistics courses. Curious about Python programming language? It will walk you through creating Python programs that parse unstructured data like language and recommends downloading Python and the Natural Language Toolkit. On a companion site, the authors have actually updated the book to work with Python 3 and NLTK 3.
Big Data Analytics Methods: Modern Analytics Techniques for the 21st Century: The Data Scientist’s Manual to Data Mining, Deep Learning & Natural Language ProcessingAuthor: Peter Ghavami
3 Ways To Make Your Own Animation
Color — Knowing which colors pair well and which combinations to avoid will help you create a more pleasant animation (or, if you’re going for dissonant imagery, create a distressing animation).
Composition — This refers to knowing how to fill the screen with your animation in an efficient, aesthetically pleasing manner.
Perspective — Knowing how to demonstrate the dimensions of an image will deepen your animation.
Anatomy — Proper understanding of anatomy can help you create structurally accurate animations. It can also help you figure out which anatomical rules you can break while maintaining a believable animation.
Script your animation. First, write down everything that you want to happen. This doesn’t just mean dialogue; you should include actions and facial expressions as well. You need to have a clear idea of exactly what’s going to happen before you start.
Draw some character sheets or make character models. What your characters will look like is also a crucial aspect of your animation film. You’ll want to have a reference to look at as you draw your frames so that the character looks consistent and real from pose to pose. Draw your characters from every different angle and with a number of different expressions. You should also draw out what they’ll be wearing, especially if what they’re wearing changes between scenes.
Make your animation sketch. This is a single drawing on a single piece of paper that shows all of the major stages of movement in a single frame. This will often result in a picture that looks like conjoined quintuplets, but it’s designed to make sure that your key frames are correctly aligned and that your motions look natural.
Try using squash and stretch. Squash and stretch is when you exaggerate motions to help the human brain perceive them as real. A common example would be when you imagine a ball. It’s much more interesting to see a ball squish down to the floor a little when it lands, instead of just seeing a normal sphere. This helps the viewer feel the motion that the ball is making.
Implementing subtle facial expressions—eyebrow movement, twitching of the corner of the mouth, and even eye (specifically pupil) movement—along with body language (e.g., standing up straight versus loosening up) will convey adequately a wide range of character emotions.
For example, if a character transitions from surprised to angry, they might start with raised eyebrows, wide eyes, and an open mouth and then transition to having a frown, narrow eyes, and bared teeth.
Draw the key frames. The key frames are the major stations of the movement that the character makes. For example, if you’re animating a character turning from left to right, the key frames would show the character facing left, then the character facing the camera, and then the character facing right.
It’s a good idea to check for flow any time you finish making a change to your document.
You’ll want to check for flow again after doing this.
Clean up the drawings. Clean any sketch lines and stray marks that distract from the movement of the character. You may even choose to ink the frames of animation, depending on what you plan to do with your work.
Advertisement
The Joy Of Remastering Your Own Linux Distro
Like most of you, I’m tired of hearing about the latest Ubuntu Linux distribution spin-off. Doesn’t it feel like everybodyhas released their very own variant of Ubuntu?
This does nothing for those working in the enterprise space, who need deep support, not one-offs. And this approach to customizing a Linux distribution isn’t all that reliable. I’m not saying there is anything wrong with Ubuntu per se, rather that the customization tools are a bit flaky to use.
Ignoring my own cynicism about customized distributions, there are circumstances where remastering Linux makes a lot of sense. The fact is, there are situations where the existing Linux solutions might not be an ideal fit. Sometimes individuals and companies are inclined to customize a Linux distribution to meet their highly specialized needs.
The reasons may vary, but generally the idea is to minimize wasted resources while maintaining complete platform control. Why install the extra software packages and code if it’s not needed?
Saving the workplace from disaster and expense
There is clear value in being able to craft a Linux server release or desktop release to meet the needs of your employer. Whether this be a large enterprise solution or perhaps something smaller, control and a customized environment present a number of hidden benefits.
What would be some of these hidden benefits? How about the desire to either save money or regain maximum control over the company’s IT department?
Due to heavy licensing fees, a push to going open source with Linux has become ever more common. So rather than end up with some blanket Linux solution for the enterprise environment, doesn’t it stand to reason that a custom Linux installation might yield a stronger benefit for everyone involved?
After all, managing extra software fluff isn’t time well spent. As always, time costs money.
In other situations the enterprise customer might be more concerned about being locked into proprietary software than any perceived cost concerns. Because companies supporting proprietary software are always free to go out of business, the affected enterprise customer would be in a world of hurt, as their data is suddenly locked away in an obsolete format.
That there’s motivation to avoid this goes without saying. No one who understands it enjoys dealing with vendor lock-in. Not unless you happen to be the proprietary vendor, that is.
By now, it should be clear that having access to an efficient, customized Linux distribution makes a lot of sense.
Remastering vs. customized solutions
How customized do you need your Linux-powered enterprise experience to be? Are we talking about graphics, software and a little bit of a lock-down with the employees?
Perhaps instead, you’re in the market for a distribution completely built from the ground up so you know that each and every piece of code passes a security audit. Whatever the need, the flexibility of customized Linux translates into solid solutions for any market or need. All that is required is the desire and the software to make it all come together.
For most enterprise situations, I believe that remastering is the way to go. It’s likely cheaper (in man hours) and simpler, as there are tools that make this easy. You also get the benefit of making changes later without much thought.
Linux from Scratch
What is “Linux From Scratch?” The idea behind Linux From Scratch is that you can work with either a base of code or instead, opt for something that is piece by piece, so you can audit all code built into the distribution. After all, if security is a concern, auditing the code by hand presents a lot of value.
The flip side to this would be that you may be going overkill with your desire to create something custom. Unless there is a security motivation, going along with a basic remastering set-up might better suit your needs.
Mastering your remastered distribution
Remember what I was saying earlier about customized Linux installations with select software, branded graphics and select enterprise-friendly restrictions put into place?
Can Science Build A Fake Wine That’s Just As Good As The Real Thing?
A new business called Replica Wines claims to produce “master forgeries” of well-known wines, deploying a throng of chemical instruments and a huge flavor database to blend near-identical versions of the wines from different grapes.
Ava Winery, meanwhile, takes a different approach, bypassing grapes entirely, and going straight for the molecules — combining flavor chemicals with ethanol and water to reproduce the experience of wine, without replicating the process.
It should be noted that it’s a bit unfair to compare the two. Replica is part of an established company — the Colorado-based Integrated Beverages Group — with a line of commercial products and a team that includes a master sommelier and several distinguished winemakers. Ava is more of a lark, a thought experiment that might never have germinated outside of the hothouse conditions of San Francisco start-up culture. Its initial offering, a bottle of imitation 1992 Dom Perignon, is not yet for sale — and may never be. But these two very different companies showcase the ambitions, and limits, of chemical analysis when it comes to subjective qualities such as flavor.
According to Replica, its patent-pending proprietary method represents the “ultimate science and wine pairing,” one that “takes the guess-work” out of making good wine. Working with Ellipse Analytics, an independent analytical chemical laboratory, the company claims to have assembled the world’s largest database of alcoholic beverage flavor profiles — essentially, a database that pairs chemical markers with sensory effects.
To duplicate a wine, laboratory staff at Ellipse analyze and quantify its characteristic “macrocomponents,” such as acidity, sugars, and tannins; and “microcomponents,” volatile flavor chemicals such as linalool and methoxypyrazine. A panel of trained tasters also produces a subjective flavor profile, a record of perceived flavor notes and intensities. The goal is to align chemistry with sensory experience, using base wines from the company’s own vineyards, as well as additional wines sourced from other California and Oregon winemakers, to make a blend that mimics the original. It’s an iterative process, with analytic laboratory’s results serving to guide the hands-on work of winemakers.
Once Replica’s blend can effectively pass for the original in sensory tests, the laboratory then confirms the accuracy of the copy. “We ensure at least ninety percent chemical similarity to the wine by which each was inspired,” promises Jaclyn Bowen, president of Ellipse Analytics.
Ninety percent sounds good, but given the exquisite sensitivity of the human sensorium, even a tiny difference, molecularly speaking, can matter quite a bit when it comes to actual flavor. So how does it go down? A tasting last week in the Popular Science offices in Manhattan placed Replica’s Pickpocket side by side with The Prisoner, the popular and critically praised California red blend that “inspired” it. Although I’m not personally a huge fan of this style of red wine, the two were comparably big-boned, juicy, and smooth-drinking, and a similar deep purple color in the glass.
Ari Walker, the company’s president, noted that there are some differences; Replica’s alcohol content, for instance, is 14.9 percent, versus the Prisoner’s 15.2 percent. Nonetheless, it was difficult for this untrained (but enthusiastic) taster to precisely pinpoint the quality of the slight distinction. Certainly, if someone had poured me a glass of one after I had quaffed a glass of the other, I would not have suspected that anything was amiss. At a retail price of $25 a bottle — about forty percent less than The Prisoner — it seems like a pretty good deal.
Walker, Replica’s president, calls this approach to winemaking “disruptive.” But is it, really? There’s nothing inherently novel about pairing science and wine; just ask Louis Pasteur. Contrary to visions of agrarian simplicity, wine is a profoundly industrialized product, and a huge global business, exceeding $250 billion dollars in annual sales. Science and technology inflects every aspect of wine production. According to Dr. Andrew Waterhouse, professor at wine chemistry at the University of California, Davis, many wineries and winemakers already try to duplicate the flavor and style of top-selling or well-reviewed wines; it’s a commercial necessity in a crowded market. There is a whole industry of wine consulting laboratories, such as Enologix and ETS Labs, which help winemakers shape their vintages to suit consumer tastes and trends. What’s unique about Replica, Waterhouse says, is simply that it’s openly admitting to the copying.
Ultimately, the “master forgery” angle is a cool trick, but it’s also a bit of misdirection. Replica’s goal, ultimately, is not to make an indistinguishable “forgery” that could fool professional wine snobs, but to consistently make wine that lots of people like, based on the wines that people are currently buying and drinking and enjoying. The big mystery here is not what makes wine taste the way it does, but what makes you like it (or, more to the point, what makes you buy it).
Despite the company’s boasts about its proprietary methodology, it’s striking how traditional some aspects of Replica’s approach and philosophy are. For instance, Walker, the company’s president, is adamant that the company would never use its methods to try to duplicate a terroir-specific wine, one made from grapes grown only on a specific parcel of land. Those wines, he suggests, have a uniqueness of character that would be impossible to replicate without access to the grapes themselves. Both he and Brett Zimmerman, the master sommelier who serves as the company’s Chief Wine Officer, seem appalled when I ask if they would ever add flavor chemicals to their wines to enhance the quality of the resemblance.
A Fruitless AttemptAva Winery is another thing altogether. In fact, it’s questionable whether it can legally call itself a winery at all, or label its product as wine. Founded last year by Mardonn Chua and Alec Lee, former college classmates with backgrounds in biotech and science education, Ava originated out of a wish to make the most exclusive wine experiences accessible to all. Seduced by an unattainable bottle of 1973 Chateau Montelena, the California Chardonnay wine that vanquished France in the famed “Judgment of Paris,” Chua and Lee wondered whether they could reproduce it chemically, molecule by molecule, without ever touching a grape. With synthetic chemistry, they could turn “water to wine in fifteen minutes.”
This is not entirely unprecedented. In the nineteenth century, synthetic chemicals were used to add distinctive flavor and specious value to neutral spirits; an 1858 editorial in Scientific American railed against chemists who peddled “flavorings to produce, at a moment’s notice, any desired liquor,” from Catawba wine to old whiskey. More recently (and more respectably), Thomas Hofmann, a flavor chemist at the Technical University of Munich, has produced mixtures of synthetic chemicals that faithfully reproduce the aroma of certain wines, using techniques that combine sensory and chemical analysis — a methodology similar to that used by Replica.
Suffice it to say, neither Chua nor Lee have any experience with winemaking, nor with flavor chemistry — although the co-founders are working with a sommelier, who has asked to remain anonymous. (“He has encountered a decent amount of disinterest in this from his colleagues,” Lee told me.)
Ava experiments with glycerin Mardonn Chua
For Ava’s “first edition,” a synthetic version of 1992 Dom Perignon, the company has sent samples of the vintage to commercial analytic laboratories, which use various forms of gas chromatography and mass spectrometry to isolate and identify its components. The challenge is interpreting those often inconsistent results: figuring out which compounds contribute to flavor, obtaining those compounds from chemical suppliers, and then recombining them in the proper quantities, a steep challenge given that some flavor compounds are used at concentrations of parts per billion or less.
So how does it taste? Reports of early versions have been less than alluring. Lisa Grossman at New Scientist detected evocative notes of plastic pool shark and raw gasoline in her recent tasting of an experimental “moscato.”
In addition to refining the formula, Ava must also clear considerable licensing and regulatory hurdles before the company can sell any alcohol. Curious quaffers can sign up for one of the 499 bottles of faux 1992 Dom ($50), but they won’t be charged for it until all of these issues are resolved.
Lee admits that this is, technically, an extremely difficult project, but insists that it’s not nearly as challenging as other ongoing attempts to produce synthetic food, such as lab-grown meat.
Lee sees Ava as part of a vanguard of technologically driven start-ups working toward a radical transformation in food production, a peer of companies such as Soylent and Impossible Foods, who have no reverence for the agrarian lore of “working the land,” and instead embrace engineering ideals such as efficiency and transparency. Indeed, Lee argues that Ava is much more transparent than the “vast majority of the wine industry,” where additives are used, but kept out of the limelight. Chua, his co-founder, published his complete formula from an earlier attempt to “hack wine” — a Kendall-Jackson chardonnay — to Reddit earlier this year. “Our customers definitely know what we’re making,” Lee says.
In this light, the company’s ambitious choice to attempt to replicate wine — a substance with profound importance throughout human history and culture — is crucial to understanding their project. “Something romantic just exists about wine itself,” Lee explains. If the company succeeds in chemically replicating such a “significant part of the cultural food chain, that’s when we really start to change people’s minds.”
Lee envisions a future where all food will be synthetic, and where our appetites will be gratified on demand by automatic replicators. “You tell your replicator what you want and then you eat it. You enjoy it. There’s nothing scary about that.” He concedes, “I might be wrong about the timescale. It might not happen in my lifetime,” yet assures me, “it’s an inevitability.” In that regard, Ava “wines” can be seen as a means of defusing the cultural and commercial landmines that persist around the word “synthetic.”
So, if you’re looking for a well-made and reliably crowd-pleasing wine at a modest price, don’t hesitate to pick up a bottle of Replica, and be sure to lift a glass to toast the science of winemaking.
But if you want to take a chance, and fully embrace the synthetic lifestyle, then sign up for an edition of Ava. It probably won’t taste like a million bucks, but it just might taste like the future.
Update the detailed information about Build Your Own Fake News Classifier With Nlp on the Kientrucdochoi.com website. We hope the article's content will meet your needs, and we will regularly update the information to provide you with the fastest and most accurate information. Have a great day!