At Lionbridge, we have deep experience helping the world's largest companies teach applications to understand audio. From virtual assistants to in-car navigation, all sound-activated machine learning systems rely on large sets of audio data. This time, we at Lionbridge combed the web and compiled this ultimate cheat sheet for public audio datasets for machine learning. AudioSet: AudioSet is an expanding ontology of 632 audio event classes and a collection of 2,084,320 human-labeled 10-second sound clips drawn from YouTube videos. LibriSpeech: LibriSpeech is a carefully segmented and aligned corpus of approximately 1000 hours of 16kHz read English speech, derived from read audiobooks. Spoken Digit Dataset: This dataset was created to solve the task of identifying spoken digits in audio samples.

At Lionbridge, we know that high quality training data can be difficult to find. To help students, data scientists, and development teams get the data they need, we've posted a large amount of dataset aggregations on our blog. Here, you can find all of those datasets in one convenient place and search for the data you need based on use case or data type. This list will be constantly updated, providing you with the best dataset aggregator available online. The datasets have been listed in alphabetical order according to use case.

Data collection is the single most important step in solving any machine learning problem. As such, teams that dive head first into projects without considering the right data collection process often don't get the results they want. Fortunately, there are many data collection tools to help prepare training datasets quickly and at scale. The best data collection tools are easy to use, support a range of functionalities and file types, and preserve the overall integrity of data. In this article, we outline the best data collection tools for machine learning projects.

To help you build object recognition models, scene recognition models, and more, we've compiled a list of the best image classification datasets. These datasets vary in scope and magnitude and can suit a variety of use cases. Furthermore, the datasets have been divided into the following categories: medical imaging, agriculture & scene recognition, and others. This goal of the competition was to use biological microscopy data to develop a model that identifies replicates. The full information regarding the competition can be found here.

With the rise of cryptocurrencies around the world, there are now more ways than ever for people to invest their money. If you could accurately predict the stock market, you'd be one of the richest people on earth. As a result, there have been previous studies on how to predict the stock market using sentiment analysis. For those of you looking to build similar predictive models, this article will introduce 10 stock market and cryptocurrency datasets for machine learning. The data was last updated on November 10th, 2017 and the files are all in CSV format.

With the emergence of crowdsourcing platforms such as Amazon Mechanical Turk, more and more companies are making crowdsourced data a key component of their machine learning strategy. By engaging a group of crowdworkers, companies can distribute hundreds of thousands of machine learning microtasks quickly and cost-effectively. Listed below are just a few of the many advantages of using crowdsourced data in machine learning. A recent study by AI market research firm Cognilytica found that nearly 80% of time spent on AI projects revolves around collecting, cleaning, and labeling data. That leaves only 20% for model development, training and calibration.

At the core of any AI project lies a great deal of annotated data for machine learning. Whether the end product is a customer service chatbot or a sentiment analysis engine, anybody building machine learning models eventually requires access to a vast amount of training data. Capturing enough accurate, quality data at scale is a common challenge for individuals and businesses alike. In this article, we outline four ways to source raw data for machine learning, and how to go about conducting data annotation. The internet contains thousands of publicly available datasets ready to be used, analyzed and enriched.

Diverse AI training data is imperative to building multilingual machine learning models, especially for morphologically complex languages like Korean. Because finding enough relevant data in Korean is difficult, we at Lionbridge have put together a comprehensive list of public Korean datasets for machine learning. National Institute of the Korean Language Corpus: This dataset contains frequency information on Korean, which is spoken by 80 million people. For each item, both the frequency (number of times it occurs in the corpus) and its relative rank to other lemmas is provided. Sentiment Lexicons for 81 Languages: This dataset contains both positive and negative sentiment lexicons for 81 languages, including Korean.

Datasets are an integral part of machine learning. Without high quality training datasets, machine learning algorithms would have no way of knowing how to conduct sentiment analysis, categorize products or understand foreign languages. This spreadsheet contains the ultimate list of open datasets for machine learning. Organized by industry and use case, this database contains a diverse range of 300 datasets to train machine learning models.

A Geographic Information System (GIS) is designed to capture, store, manipulate and present geospatial data. Machine learning is increasingly being in conjunction with GIS for a number of exciting potential benefits, such as optimizing traffic management or ride sharing applications. All location-based software are created using a large foundation of structured geospatial data. Luckily, there are a number of great sources for public geographic data free for anyone looking to build or train geographic information systems. To help, we at Lionbridge have curated a list of the 15 best publicly available geographic data sources.