This is the complete set of data files released with SCOPe 2.03-stable. All SCOPe data are described in the publications, available here: https://scop.berkeley.edu/references/ver=2.03.

Full documentation for SCOPe is available on the SCOPe website, https://scop.berkeley.edu.

Parseable files are described here: https://scop.berkeley.edu/help/ver=2.03#parseablefiles.

The SQL tables are documented here: https://scop.berkeley.edu/downloads/schema/2.03/

Context

Most supervised machine learning tasks assume a dataset with a set of well-defined target label set. But what happens when a trained model meets the real world, where inputs to the trained model might not be from the well-defined target label set? This dataset offers a way to evaluate intent classification models on "out-of-scope" inputs.

"Out-of-scope" inputs are those that do not belong to the set of "in-scope" target labels. You may have heard other ways of referring to out-of-scope, including "out-of-domain" or "out-of-distribution".

Content

• is_*.json: these files house the train/val/test sets for the in-scope data. There are 150 in-scope "intents" (aka classes), which include samples such as "what is my balance" (which belongs to the balance class).
• oos_*.json: these files house the train/val/test sets for the out-of-scope data. There is one out-of-scope intent: oos. Note that you don't have to use the oos_train.json data. In other words, an ML solution to the out-of-scope problem need not be trained on out-of-scope data, but it might help!

Evaluation Metrics

The task is intent classification, which generalizes to text classification (or categorization). This is a supervised ML problem. We use two metrics to evaluate:

• In-scope accuracy is defined as #(correctly classified in-scope samples) / #(in-scope samples).
• Out-of-scope recall is defined as #(correctly classified out-of-scope samples) / #(out-of-scope samples).

Acknowledgements

This dataset is from An Evaluation Dataset for Intent Classification and Out-of-Scope Prediction by Larson et al., which was published in EMNLP in 2019. The GitHub page for this dataset is linked here.

Inspiration

Most supervised machine learning tasks assume a dataset with a set of well-defined target label set. But what happens when a trained model meets the real world, where inputs to the trained model might not be from the well-defined target label set? This "out-of-distribution" problem has seen lots of recent development, as researchers and practitioners in both academia and industry are observing that many ML methods struggle on out-of-distribution data in a wide variety of tasks.

Despite substantial interference from hurricanes Bonnie, Charley, and Katrina, Operations Deep-Scope 2004 and 2005 were extremely successful. In addition to numerous discoveries (e.g., fluorescent sharks, new large deep-sea squid, UV vision in deep-sea crabs, the importance of polarized light and bioluminescent searchlights), these expeditions developed several new technologies. The Eye-in-the-Sea is now a robust stealth camera system; waveband, fluorescence, polarization, and UV imaging techniques are well-developed; and we can now collect deep-sea benthic species without damaging their eyes. Together these achievements place us in a unique position to explore the deep sea in innovative and exciting ways. In 2007 we propose to extend the envelope of this exciting frontier in ocean exploration in two ways: 1) developing further imaging and listening technologies, 2) using the currently developed methods to explore the cliffs in the Bahamas that range from the surface to 3000 feet in depth. Results from this proposal will characterize an important deep-sea benthic environment, and use new technologies to locate inorganic and organic ocean resources, fulfilling two of the main themes of Ocean Exploration. The proposed cliff sites range from shallow coral reefs to the abyssal plain and will allow us to explore a large number of benthic communities in a small location and learn how depth affects undersea life. Given the technological focus of our research however, if the ships' schedules should make it difficult to work in this area we would welcome the opportunity to test these new technologies at any biologically rich sites. Unlike many research cruises, which focus in depth on one problem and method, we propose a number of smaller projects that are linked by the methods and questions of visual ecology and optical oceanography. The ultimate goal of our highly interdisciplinary group of researchers is to explore and characterize the deep-sea world in these new ways.

CATH Domain Classification List (latest release) - protein structural domains classified into CATH hierarchy.

Global Credit Risk Database Market is segmented by Application (Financial Services_ Credit Scoring_ Risk Management), Type (Data Analytics_ Reporting_ Compliance Solutions), and Geography (North America_ LATAM_ West Europe_Central & Eastern Europe_ Northern Europe_ Southern Europe_ East Asia_ Southeast Asia_ South Asia_ Central Asia_ Oceania_ MEA)

Conformal Protein Retrieval - Data Files

This dataset contains the large data files required to run the Conformal Protein Retrieval Gradio Space.

  Contents





  📊 Lookup Databases

UniProt Database:

data/lookup_embeddings.npy - Pre-embedded UniProt protein sequences (Protein-Vec embeddings) data/lookup_embeddings_meta_data.tsv - Metadata for UniProt proteins (Entry, Pfam, Protein names)

SCOPE Database:

data/lookup/scope_lookup_embeddings.npy - Pre-embedded SCOPE… See the full description on the dataset page: https://huggingface.co/datasets/LoocasGoose/cpr_data.

This dataset contains water column data collected via CTD/rosette profiles taken on the SCOPE Gradients 2 cruise in the Northeast Pacific Ocean, averaged at every Niskin bottle trip. The SCOPE-Gradients program is designed to test conceptual and mathematical models of biogeochemical organization across the transition zone between different ecosystems.

Prominent rent growth indices often give strikingly different measurements of rent inflation. We create new indices from Bureau of Labor Statistics (BLS) rent microdata using a repeat-rent index methodology and show that this discrepancy is almost entirely explained by differences in rent growth for new tenants relative to the average rent growth for all tenants. Rent inflation for new tenants leads the official BLS rent inflation by four quarters. As rent is the largest component of the consumer price index, this has implications for our understanding of aggregate inflation dynamics and guiding monetary policy. Download NTRR and ATRR indices through 2022q3 here.

Dataset Description

This dataset contains a curated subset (14% of the original) of protein sequences from the Astral SCOPe 2.08 genetic domain sequence subsets.It is designed for protein family classification tasks, where the goal is to assign each amino acid sequence to its corresponding SCOPe family.

  Key Features

Source: Derived from the SCOPe database, which provides a hierarchical classification of protein structural domains based on experimental structural data… See the full description on the dataset page: https://huggingface.co/datasets/ilyass31/Protein_Family_Classification.

Global Database Audit and Protection Market is segmented by Application (Financial services_ Healthcare_ Retail_ Government_ IT and telecommunications), Type (On-premises solutions_ Cloud-based solutions_ Hybrid solutions_ Real-time monitoring tools_ Compliance management tools), and Geography (North America_ LATAM_ West Europe_Central & Eastern Europe_ Northern Europe_ Southern Europe_ East Asia_ Southeast Asia_ South Asia_ Central Asia_ Oceania_ MEA)

222 Global import shipment records of Spotting Scope with prices, volume & current Buyer's suppliers relationships based on actual Global export trade database.

Data and information collected by the submersible Johnson Sea-Link II at waypoints along its track during nineteen dives of the 2007 "Operation Deep Scope" expedition sponsored by the National Oceanic and Atmospheric Administration (NOAA) Office of Ocean Exploration, August 17 through August 27, 2007. Measurements and information include sub's position and depth; personnel assignments; dive, mission, target, and vehicle ID's; dive comments; hyperlinks to CTD plots; water temperature; and salinity. The Marine Operations Division of the Harbor Branch Oceanographic Institution provided the original submersible data.

BIOS-SCOPE BGC data 2016-2019

This ActivePaper contains the structures in version 2.04 of the ASTRAL SCOPe subset with less than 40% sequence identity. For more information about ASTRAL and SCOPe, see

http://scop.berkeley.edu/astral/

Each ASTRAL entry describes a domain from a protein structure in the PDB. This ActivePaper contains these domains in the MOSAIC HDF5 format. For more information about MOSAIC, see

http://mosaic-data-model.github.io/

The structures are arranged by its SCOPe classification. For example, ASTRAL entry d1v0aa1 is found under /data/b/18/1/30/d1v0aa1, because SCOPe classifies it as

b: all-beta proteins
b.18: Galactose-binding domain-like
b.18.1: Galactose-binding domain-like
b.18.1.30: CBM11

For each entry, the ASTRAL database provides a reference to the PDB entry with chain and residue identifiers plus a sequence. The importlet in /code/import_structures reads this information, downloads the corresponding PDB entry in mmCIF format, extracts the domain, checks that its sequence matches the one given by ASTRAL, and stores the domain in MOSAIC format.

For a small number of ASTRAL entries, this process failed for various reasons: mismatch between the ASTRAL reference and the PDB data, mismatch in the sequences, a mistake in the PDB mmCIF file, or unjustified hypotheses in the conversion script. The number of failures was deemed sufficiently small (28 failures out of 13042 entries) for not attempting a time-consuming in-detail analysis of each failure. The list of missing entries (generated automatically during the import process) can be found in this ActivePaper under /documentation/missing-entries.

Data and information collected by the submersible Johnson Sea-Link I along its track during thirteen dives of the 2005 "Operation Deep Scope" expedition sponsored by the National Oceanic and Atmospheric Administration (NOAA) Office of Ocean Exploration, August 19 through September 4, 2005. Measurements and information include sub's position, altitude, and depth; personnel assignments; dive, mission, and vehicle ID's; and sound velocity measurements. The Marine Operations Division of the Harbor Branch Oceanographic Institution provided the original submersible data.

5105 Global export shipment records of Medical Scope with prices, volume & current Buyer's suppliers relationships based on actual Global export trade database.

RDC’s Scope 3 emissions data provides CO2 emissions for each stage of flight. This innovative data is specifically designed to allow airport’s to accurately report aviation scope 3 emissions.

Scope Materials Industry Co Limited Export Import Data. Follow the Eximpedia platform for HS code, importer-exporter records, and customs shipment details.

Scope Chemicals Private Limited Export Import Data. Follow the Eximpedia platform for HS code, importer-exporter records, and customs shipment details.

5105 Global import shipment records of Medical Scope with prices, volume & current Buyer's suppliers relationships based on actual Global export trade database.