{"component": "clause", "props": {"groups": [{"samples": [{"hash": "iRoSsSKPcz3", "uri": "/contracts/iRoSsSKPcz3#machine-learning", "label": "Memorandum of Agreement", "score": 36.1937332153, "published": true}, {"hash": "lbdDZHj1ykX", "uri": "/contracts/lbdDZHj1ykX#machine-learning", "label": "Motion Picture Theatrical and Tv Series Production Agreement", "score": 36.186923354, "published": true}, {"hash": "5XmsKtpFyPD", "uri": "/contracts/5XmsKtpFyPD#machine-learning", "label": "Memorandum of Agreement", "score": 36.0189399719, "published": true}], "snippet": "The parties acknowledge that machine learning (\u2018ML\u2019) is a subset of AI that enables machines to develop algorithms, including via deep learning (as defined below), based on statistical inferences drawn from patterns in submitted training data, including, but not limited to, diffusion models and large language models, for the purpose of performing tasks. Such tasks include, but are not limited to, predicting human behaviors, disseminating information and generating content.", "size": 9, "snippet_links": [{"key": "the-parties-acknowledge", "type": "clause", "offset": [0, 23]}, {"key": "to-develop", "type": "definition", "offset": [93, 103]}, {"key": "deep-learning", "type": "clause", "offset": [130, 143]}, {"key": "based-on", "type": "definition", "offset": [164, 172]}, {"key": "training-data", "type": "definition", "offset": [229, 242]}, {"key": "not-limited", "type": "clause", "offset": [259, 270]}, {"key": "for-the-purpose-of", "type": "definition", "offset": [319, 337]}, {"key": "disseminating-information", "type": "clause", "offset": [428, 453]}], "hash": "6d48959707f9de07bd12660597841e74", "id": 1}, {"samples": [{"hash": "gYDwP2TWaDu", "uri": "/contracts/gYDwP2TWaDu#machine-learning", "label": "End User License Agreement", "score": 31.4219608307, "published": true}], "snippet": "We introduce the essential concepts and terminology related to machine learning and its application to the cybersecurity domain, and more specifically, to malware detection.\u200c", "size": 1, "snippet_links": [{"key": "related-to", "type": "definition", "offset": [52, 62]}, {"key": "its-application", "type": "definition", "offset": [84, 99]}, {"key": "more-specifically", "type": "clause", "offset": [133, 150]}], "hash": "f9c1d658cd9708c29cae71eda8e70943", "id": 7}, {"samples": [{"hash": "kGGKhGgzK8u", "uri": "/contracts/kGGKhGgzK8u#machine-learning", "label": "End User License Agreement", "score": 36.3118209839, "published": true}], "snippet": "Usage Data and Customer Content may be used to develop, train, or enhance artificial intelligence or machine learning models that are part of Provider's products and services, including third-party components of the Product, and Customer authorizes Provider to process its Usage Data and Customer Content for such purposes. However, (a) Usage Data and Customer Content must be aggregated before it can be used for these purposes, and (b) Provider will use commercially reasonable efforts consistent with industry standard technology to de-identify Usage Data and Customer Content before such use. Nothing in this section will reduce or limit Provider's obligations regarding Personal Data that may be contained in Usage Data or Customer Content under Applicable Data Protection Laws. Due to the nature of artificial intelligence and machine learning, information generated by these features may be incorrect or inaccurate. Product features that include artificial intelligence or machine learning models are not human and are not a substitute for human oversight.", "size": 2, "snippet_links": [{"key": "usage-data", "type": "clause", "offset": [0, 10]}, {"key": "customer-content", "type": "definition", "offset": [15, 31]}, {"key": "to-develop", "type": "definition", "offset": [44, 54]}, {"key": "artificial-intelligence", "type": "definition", "offset": [74, 97]}, {"key": "products-and-services", "type": "definition", "offset": [153, 174]}, {"key": "the-product", "type": "definition", "offset": [212, 223]}, {"key": "customer-authorizes", "type": "clause", "offset": [229, 248]}, {"key": "use-commercially-reasonable-efforts", "type": "clause", "offset": [452, 487]}, {"key": "consistent-with", "type": "definition", "offset": [488, 503]}, {"key": "industry-standard", "type": "definition", "offset": [504, 521]}, {"key": "nothing-in-this-section", "type": "clause", "offset": [597, 620]}, {"key": "personal-data", "type": "definition", "offset": [675, 688]}, {"key": "contained-in", "type": "definition", "offset": [701, 713]}, {"key": "applicable-data-protection-laws", "type": "definition", "offset": [751, 782]}, {"key": "nature-of", "type": "clause", "offset": [795, 804]}, {"key": "product-features", "type": "clause", "offset": [923, 939]}], "hash": "04dc4b66a47c4d270431ee1c8dbffdd2", "id": 4}, {"samples": [{"hash": "jJJEdUWyf7C", "uri": "/contracts/jJJEdUWyf7C#machine-learning", "label": "Saas and Professional Services Agreement", "score": 26.4483222961, "published": true}, {"hash": "8UJc9fpNWev", "uri": "/contracts/8UJc9fpNWev#machine-learning", "label": "Cloud Services Subscription and Professional Services Agreement", "score": 26.3182754517, "published": true}], "snippet": "Certain Subscription Content may include machine learning, which are taught and trained largely from Customer\u2019s internal data sets. Therefore, the quality of the results and outputs of the machine learning portions of the Subscription Content (such as optimized price proposals and recommended store order quantities, etc.) (\u201cMachine Learning Outputs\u201d) is heavily reliant on the quality of the Customer Data.", "size": 2, "snippet_links": [{"key": "from-customer", "type": "clause", "offset": [96, 109]}, {"key": "data-sets", "type": "definition", "offset": [121, 130]}, {"key": "quality-of-the", "type": "clause", "offset": [147, 161]}, {"key": "the-subscription", "type": "clause", "offset": [218, 234]}, {"key": "price-proposals", "type": "clause", "offset": [262, 277]}, {"key": "order-quantities", "type": "clause", "offset": [300, 316]}, {"key": "the-customer-data", "type": "clause", "offset": [390, 407]}], "hash": "cc2eccfc42b313965561089380fa2132", "id": 2}, {"samples": [{"hash": "4zxj8Qh9gnr", "uri": "/contracts/4zxj8Qh9gnr#machine-learning", "label": "Customer Agreement", "score": 36.3324317932, "published": true}, {"hash": "fJ6BujZGagh", "uri": "/contracts/fJ6BujZGagh#machine-learning", "label": "Customer Agreement", "score": 32.8144798279, "published": true}], "snippet": "Customer acknowledges that a fundamental component of the Moveworks Product is the use of machine learning for the purpose of improving and providing Moveworks\u2019 products and services. Notwithstanding anything to the contrary, Customer agrees that Moveworks is hereby granted the right to use (during and after the term hereof) IT and employee service helpdesk ticket information submitted hereunder to train its algorithms internally through machine learning techniques for such purpose.", "size": 2, "snippet_links": [{"key": "customer-acknowledges", "type": "clause", "offset": [0, 21]}, {"key": "for-the-purpose-of", "type": "definition", "offset": [107, 125]}, {"key": "products-and-services", "type": "definition", "offset": [161, 182]}, {"key": "notwithstanding-anything-to-the-contrary", "type": "clause", "offset": [184, 224]}, {"key": "customer-agrees-that", "type": "clause", "offset": [226, 246]}, {"key": "right-to-use", "type": "definition", "offset": [279, 291]}, {"key": "after-the-term", "type": "clause", "offset": [304, 318]}, {"key": "service-helpdesk", "type": "clause", "offset": [343, 359]}, {"key": "information-submitted", "type": "clause", "offset": [367, 388]}], "hash": "faf51b50cc68873a3788631da120380a", "id": 3}, {"samples": [{"hash": "uW4CxdtowU", "uri": "/contracts/uW4CxdtowU#machine-learning", "label": "Distribution Agreement", "score": 26.2950839996, "published": true}], "snippet": "Hands-On for Developers and Technical Professionals. \u2587\u2587\u2587\u2587 \u2587\u2587\u2587\u2587\u2587 & Sons. \u2587\u2587\u2587\u2587\u2587\u2587, \u2587. \u2587., \u2587\u2587\u2587\u2587\u2587\u2587, K., \u2587\u2587\u2587\u2587\u2587, J., \u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587, T., & \u2587\u2587\u2587\u2587\u2587\u2587, \u2587. \u2587. (2016). Implementation of Web-Based Autism Screening in an Urban Clinic. Clinical Pediatrics, 55(10), 927\u2013934. \u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587, \u2587. \u2587., \u2587\u2587\u2587\u2587, J., Van Naarden \u2587\u2587\u2587\u2587\u2587, K., Bilder, D., \u2587\u2587\u2587\u2587\u2587\u2587\u2587, J., \u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587, \u2587. \u2587., \u2026 Centers for Disease Control and Prevention (CDC). (2016). Prevalence and Characteristics of Autism Spectrum Disorder Among Children Aged 8 Years--Autism and Developmental Disabilities Monitoring Network, 11 Sites, United States, 2012. Morbidity and Mortality Weekly Report. Surveillance Summaries , 65(3), 1\u201323. \u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587, \u2587. \u2587., Akshoomoff, N., & \u2587\u2587\u2587\u2587\u2587\u2587\u2587, \u2587. \u2587. (2013). Diagnosis of autism spectrum disorders in 2-year-olds: a study of community practice. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 54(2), 178\u2013185. \u2587\u2587\u2587\u2587\u2587\u2587\u2587, A. M., Halladay, A. K., Shih, A., \u2587\u2587\u2587\u2587\u2587, \u2587. \u2587., & \u2587\u2587\u2587\u2587\u2587\u2587, G. (2014). Approaches to enhancing the early detection of autism spectrum disorders: a systematic review of the literature. Journal of the American Academy of Child and Adolescent Psychiatry, 53(2), 141\u2013152. \u2587\u2587\u2587\u2587 \u2587\u2587\u2587\u2587\u2587, \u2587\u2587\u2587\u2587 \u2587. \u2587\u2587\u2587\u2587, and \u2587\u2587\u2587 \u2587. \u2587\u2587\u2587\u2587\u2587\u2587. (2016). Online Appendix for \u201cData Mining: Practical Machine Learning Tools and Techniques.\u201d Data Mining: Practical Machine Learning Tools and Techniques. \u2587\u2587\u2587\u2587 \u2587\u2587\u2587\u2587\u2587\u2587\u2587. Retrieved from \u2587\u2587\u2587\u2587://\u2587\u2587\u2587.\u2587\u2587.\u2587\u2587\u2587\u2587\u2587\u2587\u2587.\u2587\u2587.\u2587\u2587/ml/weka/Witten_et_al_2016_appendix.pdf Falkmer, T., \u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587, K., Falkmer, M., & Horlin, C. (2013). Diagnostic procedures in autism spectrum disorders: a systematic literature review. European Child & Adolescent Psychiatry, 22(6), 329\u2013340. \u2587\u2587\u2587\u2587\u2587\u2587-\u2587\u2587\u2587\u2587\u2587\u2587, E., \u2587\u2587\u2587\u2587\u2587\u2587, J., & Peacock, G. (2016). Whittling Down the Wait Time: Exploring Models to Minimize the Delay from Initial Concern to Diagnosis and Treatment of Autism Spectrum Disorder. Pediatric Clinics of North America, 63(5), 851\u2013859. \u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587, \u2587. \u2587., Bai, R., & \u2587\u2587\u2587\u2587\u2587\u2587\u2587, A. M. (2013). Screening children for autism in an urban clinic using an electronic M-CHAT. Clinical Pediatrics, 52(1), 35\u201341. How Is Autism Diagnosed? (n.d.). Retrieved February 8, 2017, from \u2587\u2587\u2587\u2587\u2587://\u2587\u2587\u2587.\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587.\u2587\u2587\u2587/what-autism/diagnosis \u2587\u2587\u2587\u2587\u2587\u2587\u2587, \u2587. \u2587., & \u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587, \u2587. \u2587. (2004). Evaluation of reporting timeliness of public health surveillance systems for infectious diseases. BMC Public Health, 4, 29. \u2587\u2587\u2587\u2587\u2587, M. (2017). CDC: Autism rates unchanged at 1 in 68 children. AAP News. Retrieved from \u2587\u2587\u2587\u2587://\u2587\u2587\u2587.\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587.\u2587\u2587\u2587/news/2016/03/31/AutismRates033116 \u2587\u2587\u2587, M., \u2587\u2587\u2587\u2587\u2587\u2587, J., & Takane, M. (2011). mHealth: New horizons for health through mobile technologies. World Health Organization, 3, 66\u201371. Klin, A., \u2587\u2587\u2587\u2587\u2587\u2587\u2587, C., & \u2587\u2587\u2587\u2587\u2587, W. (2015). Reducing age of autism diagnosis: developmental social neuroscience meets public health challenge. Revista de Neurologia, 60 Suppl 1, S3\u2013 11. \u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587, \u2587. \u2587., Sochat, V., \u2587\u2587\u2587\u2587, M., & Wall, D. P. (2015). Searching for a minimal set of behaviors for autism detection through feature selection-based machine learning. Translational Psychiatry, 5, e514. \u2587\u2587\u2587\u2587, R., & Colditz, G. A. (2013). Implementation science and its application to population health. Annual Review of Public Health, 34, 235\u2013251.", "size": 1, "snippet_links": [{"key": "centers-for-disease-control-and-prevention", "type": "definition", "offset": [351, 393]}, {"key": "characteristics-of", "type": "definition", "offset": [424, 442]}, {"key": "developmental-disabilities", "type": "clause", "offset": [508, 534]}, {"key": "monitoring-network", "type": "clause", "offset": [535, 553]}, {"key": "united-states", "type": "clause", "offset": [565, 578]}, {"key": "weekly-report", "type": "definition", "offset": [610, 623]}, {"key": "diagnosis-of-autism-spectrum-disorders", "type": "definition", "offset": [721, 759]}, {"key": "review-of-the-literature", "type": "clause", "offset": [1056, 1080]}, {"key": "data-mining", "type": "clause", "offset": [1240, 1251]}, {"key": "diagnostic-procedures", "type": "definition", "offset": [1522, 1543]}, {"key": "systematic-literature-review", "type": "clause", "offset": [1576, 1604]}, {"key": "wait-time", "type": "clause", "offset": [1734, 1743]}, {"key": "treatment-of", "type": "clause", "offset": [1822, 1834]}, {"key": "north-america", "type": "clause", "offset": [1882, 1895]}, {"key": "an-electronic", "type": "clause", "offset": [2021, 2034]}, {"key": "reporting-timeliness", "type": "clause", "offset": [2253, 2273]}, {"key": "surveillance-systems", "type": "clause", "offset": [2291, 2311]}, {"key": "infectious-diseases", "type": "definition", "offset": [2316, 2335]}, {"key": "new-horizons", "type": "definition", "offset": [2571, 2583]}, {"key": "mobile-technologies", "type": "clause", "offset": [2603, 2622]}, {"key": "world-health-organization", "type": "definition", "offset": [2624, 2649]}, {"key": "its-application", "type": "definition", "offset": [3116, 3131]}, {"key": "population-health", "type": "definition", "offset": [3135, 3152]}, {"key": "annual-review", "type": "definition", "offset": [3154, 3167]}], "hash": "d135700b618a1cb3f97908f4eb106e95", "id": 5}, {"samples": [{"hash": "jhW9XFiim0d", "uri": "/contracts/jhW9XFiim0d#machine-learning", "label": "Terms of Service", "score": 34.2099342346, "published": true}], "snippet": "Client agrees and instructs that Heyday may use Conversational Data to develop and improve the capabilities of the Services and Heyday's machine-learning technologies, both during and after the term of the Agreement, provided that (i) such Conversational Data shall be anonymized so that no individual can be specifically identified; and (ii) such Conversational Data shall not be shared with any other client or customer.", "size": 1, "snippet_links": [{"key": "client-agrees", "type": "clause", "offset": [0, 13]}, {"key": "to-develop", "type": "definition", "offset": [68, 78]}, {"key": "services-and", "type": "clause", "offset": [115, 127]}, {"key": "the-term-of-the-agreement", "type": "clause", "offset": [190, 215]}, {"key": "provided-that", "type": "definition", "offset": [217, 230]}, {"key": "client-or-customer", "type": "definition", "offset": [403, 421]}], "hash": "0864a0f005bca337c3e531f5b5ca11ee", "id": 9}, {"samples": [{"hash": "bhJIbvlkNX9", "uri": "/contracts/bhJIbvlkNX9#machine-learning", "label": "Distribution Agreement", "score": 28.8945674896, "published": true}], "snippet": "Machine learning is the process of learning patterns from available data to make predictions that generalize to \u201cfuture unseen\u201d data. It is generally divided into two major types: supervised and unsupervised learning. When labels are available for the dataset, a supervised learning approach is often used to learn how to predict these labels from the features provided. When labels are not available, an \u201cunsupervised\u201d approach is used, where there is no phenotype or outcome to predict, but a supposed underlying structure of the data is being discovered. A fundamental concept in machine learning is data separation and the quest for generalization. Before making any predictions, the data is divided into training and testing sets. The testing set, also known as the \u201cheld out\u201d set, is used to test how generalizable the trained model would be if it were to be used on future unseen data. A simple example to illustrate this is polynomial fitting. Suppose we have two synthetically-generated random variables (X1 and X2), which when drawn in a scatter plot (with X1 and X2 being the two axes) have no underlying pattern. Given a polynomial fitting algorithm, it is possible to explain much of the variance of the data with a very high-degree polynomial, given enough training iterations. In other words, without restraint on model complexity, it is possible to explain almost any dataset to an arbitrary level of accuracy. This does not mean, of course, that the model will have any meaning or generalization, and indeed our high degree polynomial is very unlikely to be even close to accurate when it is applied to the testing data. This is known as model \u201coverfitting\u201d. There is a well-known trade-off between model fitting and generalization, and there almost always exists a \u201csweet spot\u201d where the model fits the training data well enough to have any meaning, but is generalizable enough to allow for utility over future unseen data. Most of the machines learning algorithms require tuning of model \u201chyperparameters.\u201d In regularized linear models, for example, it is necessary to determine how much to penalize the weights, and in neural networks, it is necessary to determine what network architecture and learning behavior, including the number of nodes per layer (width), the number of layers (depth), the learning rate, the type of non-linearity and the type of optimizer to use. If we were to tune these parameters on the testing set, we would be defeating the purpose of an independent, untouched testing set to evaluate model generalization. Therefore, the training set is often further divided into training and validation sets, where the validation set is used to provide feedback on model generalizability to choose the most optimum hyperparameter configuration. We are proposing the use of a Neural Networks machine learning approach to train the matching algorithm. This choice is based on recent successes neural networks have had in handing some of the most difficult prediction tasks, especially in the image classification, natural language processing and time series fields. Machine learning using neural networks with more than one layer is often called Deep Learning, and has exploded in popularity over the last century because of the advent of novel analytic/computational approaches as well as the increased power of computing and availability of \u201cBig Data.\u201d Originally modeled to mimic the human Central Nervous System (CNS) and later taking on a life of its own, neural networks are fundamentally very simple. Each node represents the result of a linear combination of input features, with different weights and added biases, followed by the application of a non-linearity (known as the activation function). The most common nonlinearities used are sigmoidal, Tanh and Rectified Linear Unit (ReLU). Logistic regression is, essentially, a single node with a sigmoidal activation function (also known as a perceptron). The power of neural networks comes from combining multiple perceptrons in clever ways horizontally (forming nodes of a single layer) and vertically (forming multiple layers) to approximate functions to an arbitrary level of accuracy. The key algorithm that resulted in the explosive rise of deep learning and that made neural-based learning mainstream is backpropagation, the process of using the multivariate chain-rule to update all parameters (i.e. all weights and biases) simultaneously after each pass through the network, thus saving considerable time and effort (18,19). Multiple reasons exist for our preference of using neural networks (with NLP) in this project. Our algorithm will superimpose neural network modeling on textural NLP. The approach provides two strong artificial intelligence tools that synergize each other. The model is expected to extract textual data from \u2587\u2587\u2587.\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587\u2587.\u2587\u2587\u2587, which given", "size": 1, "snippet_links": [{"key": "the-process", "type": "clause", "offset": [20, 31]}, {"key": "available-data", "type": "clause", "offset": [58, 72]}, {"key": "the-dataset", "type": "clause", "offset": [248, 259]}, {"key": "supervised-learning", "type": "definition", "offset": [263, 282]}, {"key": "not-available", "type": "clause", "offset": [387, 400]}, {"key": "structure-of-the", "type": "clause", "offset": [515, 531]}, {"key": "data-separation", "type": "clause", "offset": [603, 618]}, {"key": "training-and-testing", "type": "clause", "offset": [709, 729]}, {"key": "the-testing", "type": "definition", "offset": [736, 747]}, {"key": "trained-model", "type": "definition", "offset": [825, 838]}, {"key": "very-high", "type": "definition", "offset": [1229, 1238]}, {"key": "the-model", "type": "clause", "offset": [1463, 1472]}, {"key": "close-to", "type": "definition", "offset": [1580, 1588]}, {"key": "testing-data", "type": "clause", "offset": [1624, 1636]}, {"key": "training-data", "type": "definition", "offset": [1821, 1834]}, {"key": "for-utility", "type": "clause", "offset": [1905, 1916]}, {"key": "for-example", "type": "clause", "offset": [2056, 2067]}, {"key": "to-determine", "type": "clause", "offset": [2085, 2097]}, {"key": "network-architecture", "type": "clause", "offset": [2190, 2210]}, {"key": "number-of", "type": "clause", "offset": [2248, 2257]}, {"key": "type-of", "type": "definition", "offset": [2336, 2343]}, {"key": "purpose-of", "type": "clause", "offset": [2474, 2484]}, {"key": "an-independent", "type": "clause", "offset": [2485, 2499]}, {"key": "to-provide", "type": "clause", "offset": [2678, 2688]}, {"key": "based-on", "type": "definition", "offset": [2901, 2909]}, {"key": "deep-learning", "type": "clause", "offset": [3180, 3193]}, {"key": "power-of", "type": "clause", "offset": [3338, 3346]}, {"key": "availability-of", "type": "clause", "offset": [3361, 3376]}, {"key": "big-data", "type": "definition", "offset": [3378, 3386]}, {"key": "the-human", "type": "clause", "offset": [3417, 3426]}, {"key": "central-nervous-system", "type": "definition", "offset": [3427, 3449]}, {"key": "application-of", "type": "clause", "offset": [3674, 3688]}, {"key": "a-non", "type": "clause", "offset": [3689, 3694]}, {"key": "logistic-regression", "type": "clause", "offset": [3831, 3850]}, {"key": "all-parameters", "type": "definition", "offset": [4380, 4394]}, {"key": "pass-through", "type": "definition", "offset": [4451, 4463]}, {"key": "the-network", "type": "clause", "offset": [4464, 4475]}, {"key": "time-and-effort", "type": "clause", "offset": [4502, 4517]}, {"key": "preference-of", "type": "clause", "offset": [4558, 4571]}, {"key": "neural-network", "type": "definition", "offset": [4653, 4667]}, {"key": "the-approach", "type": "clause", "offset": [4694, 4706]}, {"key": "artificial-intelligence", "type": "definition", "offset": [4727, 4750]}], "hash": "45f4ad171a5a3876706b13f54f3dfb8e", "id": 6}, {"samples": [{"hash": "bG85oukbr5G", "uri": "/contracts/bG85oukbr5G#machine-learning", "label": "Not Applicable", "score": 22.7152633667, "published": true}], "snippet": "Random Forest Machine Learning \u2013 Neural Networks", "size": 1, "snippet_links": [{"key": "random-forest", "type": "clause", "offset": [0, 13]}], "hash": "9ee89c20ced4520e33e3f9145ab36c92", "id": 8}, {"samples": [{"hash": "4Hn45unMujN", "uri": "/contracts/4Hn45unMujN#machine-learning", "label": "End User License Agreement", "score": 27.3723468781, "published": true}], "snippet": "the End-User is entitled to access or use the Processing Blocks for the purpose of developing or training machine learning algorithms.", "size": 1, "snippet_links": [{"key": "access-or-use", "type": "clause", "offset": [28, 41]}, {"key": "the-processing", "type": "clause", "offset": [42, 56]}, {"key": "for-the-purpose-of", "type": "definition", "offset": [64, 82]}], "hash": "6174a61b8342f12569805f35ac6a2b77", "id": 10}], "next_curs": "ClkSU2oVc35sYXdpbnNpZGVyY29udHJhY3RzcjULEhZDbGF1c2VTbmlwcGV0R3JvdXBfdjU2IhltYWNoaW5lLWxlYXJuaW5nIzAwMDAwMDBhDKIBAmVuGAAgAA==", "clause": {"children": [["machine-learning", "Machine Learning"], ["cooperation", "Cooperation"], ["prediction-based-analysis-in-gene-expression-experiments", "Prediction Based Analysis in Gene Expression Experiments"], ["machine-learning-tasks", "Machine Learning Tasks"], ["evaluating-ml-classifiers", "Evaluating ML Classifiers"]], "title": "Machine Learning", "size": 29, "parents": [["definitions", "Definitions"], ["artificial-intelligence", "Artificial Intelligence"], ["payment-of-wages", "PAYMENT OF WAGES"], ["work-in-extreme-heat-and-cold", "Work in Extreme Heat and Cold"], ["machine-learning", "Machine Learning"]], "id": "machine-learning", "related": [["service-monitoring-analyses-and-oracle-software", "SERVICE MONITORING, ANALYSES AND ORACLE SOFTWARE", "SERVICE MONITORING, ANALYSES AND ORACLE SOFTWARE"], ["computer", "Computer", "Computer"], ["evaluation-software", "Evaluation Software", "Evaluation Software"], ["workstation-laptop-encryption", "Workstation/Laptop encryption", "Workstation/Laptop encryption"], ["games", "Games", "Games"]], "related_snippets": [], "updated": "2025-07-24T04:27:57+00:00", "also_ask": ["What safeguards should be included to address data privacy and IP ownership in machine learning clauses?", "How can liability for algorithmic errors or bias be allocated or limited in negotiations?", "What are the key differences in enforceability of machine learning clauses across major jurisdictions?", "Which performance metrics or audit rights are essential for effective oversight of machine learning systems?", "How have courts interpreted obligations or warranties related to machine learning outputs in recent cases?"], "drafting_tip": "Define 'Machine Learning' precisely to avoid ambiguity; specify permitted uses to ensure compliance; address data ownership to prevent disputes.", "explanation": "A Machine Learning clause defines the terms under which machine learning technologies, data, or outputs are used, shared, or developed within the context of an agreement. It typically addresses issues such as ownership of training data, rights to use machine learning models, and responsibilities for accuracy or bias in outputs. For example, it may specify whether one party can use data provided by another to train algorithms, or who retains rights to improvements made through machine learning processes. The core function of this clause is to clarify intellectual property rights and responsibilities related to machine learning, thereby reducing disputes and ensuring both parties understand how such technologies and data can be utilized."}, "json": true, "cursor": ""}}