Artificial Intelligence for Language Translation?
Researchers are using artificial intelligence in new, creative ways. Some use cases – predicting earthquakes, recognizing objects, and comparing medical scans, for example – offer tantalizing glimpses of a promising future. Language translation has long been considered a natural application for AI, but processes are ripe for improvement. Now, MIT researchers have announced an “unsupervised” language translation model whose early returns signal a future with “faster, more efficient computer-based translations of far more languages.”
Most widely-used translation programs (like Google’s) learn via models. These programs are trained to “look for patterns in millions of documents – such as legal and political documents, or news articles” that have been translated from one language to another by humans. Once they identify a new word in one language, they can find its match in a different language through pattern analysis.
This approach, however, is labor intensive and inefficient – partially due to its reliance on specific translations from one language to another to accurately translate words. New “monolingual” models have attempted to rectify these issues by translating “without direct translational information between the two,” but are slow and require significant computing power to work.
To combat these pain points, MIT researchers with the Computer Science and Artificial Intelligence Laboratory (CSAIL) developed a new technique. By employing a statistical metric commonly used for pixel alignment called the Gromov-Wasserstein distance – which “essentially measures distances between points in one computational space and matches them to similarly distanced points in another space” – in conjunction with a vectorized system called “word embeddings”, where “words of similar meanings [are] clustered closer together,” researchers were able to create a system that can deduce likely direct translations via the relative distances of words within each vector.
Using relational distances negates the time-consuming, laborious process of creating perfect word alignments. Gromov-Wasserstein is “tailor-made” for this purpose, said CSAIL Ph.D. student David Alvarez-Melis, who was the first author of the paper presenting the findings. “If there are points, or words, that are close together in one space, Gromov-Wasserstein is automatically going to try to find the corresponding cluster of points in the other space,” explained Alvarez-Melis.
Researchers are discovering innovative ways to utilize artificial intelligence. Applications such as earthquake prediction, object recognition, and medical scan comparisons show the potential for a promising future. While AI has long been employed for language translation, improvements are still necessary. MIT researchers have introduced an "unsupervised" language translation model that has demonstrated promising results, paving the way for faster and more efficient translations of a wider range of languages.
Commonly used translation programs, such as Google's, rely on models that learn by analyzing millions of translated documents, such as legal and political papers, and news articles. These models detect patterns to locate corresponding words in different languages when encountering a new word in one language.
While traditional approaches to translation rely on specific translations from one language to another, they can be labor-intensive and inefficient. Although new "monolingual" models attempt to solve this problem by translating "without direct translational information between the two," these models are often slow and require significant computing power.
To address these challenges, researchers from MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) developed a new technique. They combined the Gromov-Wasserstein distance, a statistical metric commonly used for pixel alignment, with a vectorized system called "word embeddings." This system clusters words with similar meanings closer together, allowing researchers to deduce likely direct translations based on the relative distances of words within each vector.
Using relational distances eliminates the need for the time-consuming and laborious process of creating perfect word alignments. According to CSAIL Ph.D. student David Alvarez-Melis, the Gromov-Wasserstein distance is "tailor-made" for this purpose. He explains that if there are points or words that are close together in one space, the Gromov-Wasserstein distance will automatically try to find the corresponding cluster of points in the other space. For example, despite differences between languages, the model could identify a cluster of 12 vectors for the months of the year in one embedding and a similar cluster in the other embedding, allowing for simultaneous alignment of an entire vector space.
The result is a system with accuracy comparable to existing monolingual models but with greater speed and significantly less operating power. This is a significant step towards the goal of achieving truly unsupervised word alignment and demonstrates the power of AI being fine-tuned and deployed in new, creative, and useful ways.
If You’re Wondering When A.I. Will Start Making Market Predictions…
Did you know that Artificial Intelligence (A.I.) is already being used to make market predictions? Hedge funds and large institutional investors have been using A.I. to analyze vast amounts of data for investment opportunities and to identify patterns and trends in charts. The A.I. can scan thousands of securities and cryptocurrencies and generate trade ideas based on what it finds. However, retail investors have been at a disadvantage until now.
Tickeron has launched a new investment platform that provides retail investors with access to sophisticated A.I. for a variety of functions,
- including finding stock patterns and trends,
- testing portfolio diversification,
- back-testing trading results based on different stock patterns,
- and making predictions about future price movements with an "A.I. Rank" and level of confidence in the trade.
With Tickeron's platform, retail investors can level the playing field and take advantage of A.I. technology to make better investment decisions.
And much more. No longer is AI just confined to the biggest hedge funds in the world. It can now be accessed by everyday investors. Learn how on Tickeron.com.
Tickeron’s Approach to Fintech: Artificial Intelligence for Retail Investors
Hedge funds and large institutional investors have been using Artificial Intelligence to analyze large data sets for investment opportunities, and they have also unleashed A.I. on charts to discover patterns and trends. Not only can the A.I. scan thousands of individual securities and cryptocurrencies for patterns and trends, and it generates trade ideas based on what it finds. Hedge funds have had a leg-up on the retail investor for some time now.
Not anymore. Tickeron has launched a new investment platform, and it is designed to give retail investors access to sophisticated AI for a multitude of functions:
Finding stock patterns in the market
Finding trends in the stock market
Testing portfolios to see if they are well-diversified
Back-testing statistics to see how different stock patterns generated trading results
Making Predictions for price movements in the future, with “A.I. Rank” and level of confidence in the trade.
And much more. No longer is AI just confined to the biggest hedge funds in the world. It can now be accessed by everyday investors. Learn how on Tickeron.com.
Contributor
Sergey Savastiouk, Ph.D. has a degree in Applied Mathematics from Moscow University and has extensive experience as an entrepreneur, investor, manager, and mathematician. His professional expertise is in applied mathematics, mathematical modeling, system and pattern analysis, and software and hardware system integration. He has served as the CEO of several hi-tech start-up companies and nonprofit organizations, which has given him proven capabilities in business strategy for high-tech start-up companies, market assessment, company formation, team building, product development, marketing, and sales. He has published numerous articles in journals and magazines on related fields. As a retail investor, he spent 15 years developing his proprietary trading and quantitative algorithms (now Tickeron’s A.I.), which brought him significant returns in trading the stock market. His current work and goal in founding Tickeron is to bring professional, sophisticated stock market analysis capabilities to retail investors via an easy-to-use interface.
Advertisement