Dataset Card for Audio Keyword Spotting

  Dataset Summary

The initial version of this dataset is a subset of MLCommons/ml_spoken_words, which is derived from Common Voice, designed for easier loading. Specifically, the subset consists of ml_spoken_words files filtered by the names and placenames transliterated in Bible translations, as found in trabina. For our initial experiment, we have focused only on English, Spanish, and Indonesian, three languages whose name… See the full description on the dataset page: https://huggingface.co/datasets/sil-ai/audio-keyword-spotting.

You can check the fields description in the documentation: current Keyword database: https://docs.dataforseo.com/v3/databases/google/keywords/?bash; Historical Keyword database: https://docs.dataforseo.com/v3/databases/google/history/keywords/?bash. You don’t have to download fresh data dumps in JSON or CSV – we can deliver data straight to your storage or database. We send terrabytes of data to dozens of customers every month using Amazon S3, Google Cloud Storage, Microsoft Azure Blob, Eleasticsearch, and Google Big Query. Let us know if you’d like to get your data to any other storage or database.

OAGKX is a keyword extraction/generation dataset consisting of 22674436 abstracts, titles and keyword strings from scientific articles. The texts were lowercased and tokenized with Stanford CoreNLP tokenizer. No other preprocessing steps were applied in this release version. Dataset records (samples) are stored as JSON lines in each text file.

The data is derived from OAG data collection (https://aminer.org/open-academic-graph) which was released under ODC-BY license.

This data (OAGKX Keyword Generation Dataset) is released under CC-BY license (https://creativecommons.org/licenses/by/4.0/).

If using it, please cite the following paper:

Çano Erion, Bojar Ondřej. Keyphrase Generation: A Multi-Aspect Survey. FRUCT 2019, Proceedings of the 25th Conference of the Open Innovations Association FRUCT, Helsinki, Finland, Nov. 2019

To reproduce the experiments in the above paper, you can use the first 100000 lines of part_0_0.txt file.

This dataset consists of keywords extracted from the LibriSpeech recordings, designed to provide conversational keywords for keyword recognition tasks. It serves a similar purpose to the Speech Commands dataset.

The keywords in both the /train/ and /test/ match those found in the Speech Commands /train/ set. As a result, the /test/ set contains more keyword classes than the Speech Commands /test/ set. If a keyword is missing from either /train/ or /test/, this means that it was not found in the LibriSpeech data.

The _background_noise_ folder includes the same noise samples used in the Speech Commands dataset.

The keywords were extracted using the Montreal Force Aligner. In certain cases identified as incorrect keyword segmentations, manual correction was applied.

The "Famous Keyword Twitter Replies Dataset" is a comprehensive collection of Twitter data that focuses on popular keywords and their associated replies. This dataset contains five essential columns that provide valuable insights into the Twitter conversation dynamics:

1. Keyword: This column represents the specific keyword or topic of interest that generated the original tweet. It helps identify the context or subject matter around which the conversation revolves.

2. Main_tweet: The main_tweet column contains the original tweet related to the keyword. It serves as the starting point or focal point of the conversation and often provides essential information or opinions on the given topic.

3. Main_likes: This column provides the number of likes received by the main_tweet. Likes serve as a measure of engagement and indicate the level of popularity or resonance of the original tweet within the Twitter community.

4. Reply: The reply column consists of the replies or responses to the main_tweet. These replies may include comments, opinions, additional information, or discussions related to the keyword or the original tweet itself. The replies help capture the diverse perspectives and conversations that emerge in response to the main_tweet.

5. Reply_likes: This column records the number of likes received by each reply. Similar to the main_likes column, the reply_likes column measures the level of engagement and popularity of individual replies. It enables the identification of particularly noteworthy or well-received replies within the dataset.

By analyzing this "Famous Keyword Twitter Replies Dataset," researchers, analysts, and data scientists can gain valuable insights into how popular keywords spark discussions on Twitter and how these discussions evolve through replies.

The dataset's information on likes allows for the evaluation of tweet and reply popularity, helping to identify influential or impactful content.

This dataset serves as a valuable resource for various applications, including sentiment analysis, trend identification, opinion mining, and understanding social media dynamics.

Number of tweets for each pairs of tweet and reply

Total has 17255 pairs of tweet/reply

https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F9998584%2Fc33bf662ec0b710877ba40287bc6025e%2Fcount.png?generation=1686152411950305&alt=media" alt="">

Between June 2022 and March 2023, the traffic volume for the keyword "AI" has tripled, going from around 7.9 million monthly searches to more than 30.4 million during the last month of the measured period. General interest in artificial intelligence (AI) has exploded in markets like the United States by the end of 2022. Likewise, interest for the application programming interfaces (API's) and plugins of artificial intelligence solutions, especially those of ChatGPT, has also seen a major increase since the release of the tool in November of 2022.

The artificial intelligence market

Valued at around 142.3 billion U.S. dollars in 2022, the artificial intelligence market is one the most promising tech segments for the rest of the decade, with more than five billion U.S. dollars invested in startups - the most notable being the Californian company OpenAI and its flagship application ChatGPT. Disruptive as it is, the adoption of AI has already sparked an alert for several industries, likely to affect job markets and thus raising concerns about cybercrime and other online misdeeds.

The future of online search?

Of most industries, the impact of the new tool developed by OpenAI may be felt by the online search market like a global earthquake. With chatbots providing search results in a dialogue format, the trend of AI-powered search engines unleashed by ChatGPT threw giant companies like Google and Microsoft into a race with startups and other competitors to present the best candidate for this disruptive (and experimental) online solution.

Keyword Performance Dataset Overview

Gathered over 7 months of a worldwide Google Ads campaign targeting keywords related to AI software. The dataset provides insights into search terms that triggered ads, their performance, and associated costs. Each row in the dataset corresponds to a unique search term and its respective metrics. Below is a description of each column in the dataset:

1. Search term: The specific query entered by a user that triggered the ad. This reflects actual user intent and can reveal new opportunities for targeting or refining keywords.

2. Match type: Indicates how closely the search term matches the targeted keyword. Common match types include:

   • Exact Match: The search term matches the keyword exactly.
   • Phrase Match: The search term contains the keyword or a close variation in sequence.
   • Broad Match: The search term is loosely related to the keyword, providing a wider reach.

3. Impr. (Impressions): The number of times the ad was displayed to users. This metric helps gauge the visibility of the ad for each search term.

4. Clicks: The number of times users clicked on the ad after seeing it. This measures engagement and relevance of the ad to the search term.

5. Currency code: The currency in which the campaign costs are reported (e.g., USD, EUR). It ensures financial metrics like cost-per-click (CPC) are appropriately understood.

6. Avg. CPC (Average Cost-Per-Click): The average amount paid for each click on the ad triggered by the search term. It provides insights into the cost-efficiency of the campaign.

7. Keyword: The targeted keyword that matched the search term. Understanding the relationship between the keyword and the search term can inform optimizations, such as adding negative keywords or refining match types.

This dataset provides a foundation for analyzing campaign performance, identifying trends, and optimizing ad spend. By exploring metrics like impressions, clicks, and cost-per-click, advertisers can refine targeting and improve ROI.

Wikipedia Paragraph and Keyword Dataset

  Dataset Summary

This dataset contains 10,693 paragraphs extracted from English Wikipedia articles, along with corresponding search-engine style keywords for each paragraph. It is designed to support tasks such as text summarization, keyword extraction, and information retrieval.

  Dataset Structure

The dataset is structured as a collection of JSON objects, each representing a single paragraph with its associated keywords.… See the full description on the dataset page: https://huggingface.co/datasets/agentlans/wikipedia-paragraph-keywords.

HENSOLDT ANALYTICS MediaMiningIndexer KWS - keyword spotting engine that provides conversion of speech to text. It returns list of keywords found in the lattice of the spoken sentences. Service is provided in a variety of languages.

Datasets for evaluation of keyword extraction in Russian

The Qualcomm Keyword Speech Dataset consists of 4270 utterances belonging to four classes, with variable durations from 0.48s to 1.92s.

"Squid Game" was the keyword that had the highest maximum monthly search volume in 2021, with 101 million online searches worldwide in its peak month. The second most trending keyword based on its peak search volume was "Christian Eriksen." Eriksen is a Danish football player who experienced cardiac arrest during the Euro 2020 Denmark against Finland match, on June 12, 2021. Another popular keyword search on Google was Queen Elisabeth's late husband "Prince Phillip," with a peak of 37.2 million searches. Another popular topic related to the British royal family was Prince Harry and Meghan Markle's interview with Oprah Winfrey in March 2021. This was the first big interview after the couple decided to step back as senior royals, and queries on the topic went up to 1.2 million searches on Google in 2021.

Previous studies on supporting free-form keyword queries over RDBMSs provide users with linked-structures (e.g.,a set of joined tuples) that are relevant to a given keyword query. Most of them focus on ranking individual tuples from one table or joins of multiple tables containing a set of keywords. In this paper, we study the problem of keyword search in a data cube with text-rich dimension(s) (so-called text cube). The text cube is built on a multidimensional text database, where each row is associated with some text data (a document) and other structural dimensions (attributes). A cell in the text cube aggregates a set of documents with matching attribute values in a subset of dimensions. We define a keyword-based query language and an IR-style relevance model for coring/ranking cells in the text cube. Given a keyword query, our goal is to find the top-k most relevant cells. We propose four approaches, inverted-index one-scan, document sorted-scan, bottom-up dynamic programming, and search-space ordering. The search-space ordering algorithm explores only a small portion of the text cube for finding the top-k answers, and enables early termination. Extensive experimental studies are conducted to verify the effectiveness and efficiency of the proposed approaches. Citation: B. Ding, B. Zhao, C. X. Lin, J. Han, C. Zhai, A. N. Srivastava, and N. C. Oza, “Efficient Keyword-Based Search for Top-K Cells in Text Cube,” IEEE Transactions on Knowledge and Data Engineering, 2011.

This dataset consists of keywords extracted from the LibriSpeech recordings. It is designed to provide conversational keywords for keyword recognition tasks and serves a similar purpose to the Speech Commands dataset. Because the dataset includes data derived from LibriSpeech, users must comply with the original LibriSpeech license when using this dataset.

The keywords in both the /train/ and /test/ match those found in the Speech Commands /train/ set. As a result, the /test/ set contains more keyword classes than the Speech Commands /test/ set. If a keyword is missing from either /train/ or /test/, this means that it was not found in the LibriSpeech data.

The _background_noise_ folder includes the same noise samples used in the Speech Commands dataset.

The keywords were extracted using the Montreal Force Aligner. In certain cases identified as incorrect keyword segmentations, manual correction was applied.

umuth/keyword dataset hosted on Hugging Face and contributed by the HF Datasets community

KEYWORD SEARCH IN TEXT CUBE: FINDING TOP-K RELEVANT CELLS BOLIN DING, YINTAO YU, BO ZHAO, CINDY XIDE LIN, JIAWEI HAN, AND CHENGXIANG ZHAI Abstract. We study the problem of keyword search in a data cube with text-rich dimension(s) (so-called text cube). The text cube is built on a multidimensional text database, where each row is associated with some text data (e.g., a document) and other structural dimensions (attributes). A cell in the text cube aggregates a set of documents with matching attribute values in a subset of dimensions. A cell document is the concatenation of all documents in a cell. Given a keyword query, our goal is to find the top-k most relevant cells (ranked according to the relevance scores of cell documents w.r.t. the given query) in the text cube. We define a keyword-based query language and apply IR-style relevance model for scoring and ranking cell documents in the text cube. We propose two efficient approaches to find the top-k answers. The proposed approaches support a general class of IR-style relevance scoring formulas that satisfy certain basic and common properties. One of them uses more time for pre-processing and less time for answering online queries; and the other one is more efficient in pre-processing and consumes more time for online queries. Experimental studies on the ASRS dataset are conducted to verify the efficiency and effectiveness of the proposed approaches.

This file contains total hits per keyword expressed as percentage of total hits for the eight domains of the human well-being index. Additional categorical data is given for each community planning document based on publicly available demographic data for the community. These demographic data include population size, proportion of population in a series of categories: education level, median income, and race. Additional categorical variables are community assignment based on a community typology. A full description of the community typology can be found in the associated supplementary material. This dataset is associated with the following publication: Fulford, R., M. Russell, J. Harvey, and M. Harwell. Sustainability at the community level: Searching for common ground as a part of a national strategy for decision support. U.S. Environmental Protection Agency, Washington, DC, USA, 2016.

Dataset

This dataset was created by Maunish dave

Contents

The dataset consists of 7514 Slovenian news articles from the SentiNews 1.0 corpus by Bučar et al. 2017 (http://hdl.handle.net/11356/1110) which had available article keywords. We provide the train and test data splits (5995 articles for training and 1519 for testing) that can be used for keyword extraction experiments. The format is a json file, containing the following fields: title, keywords, lang (always Slovene) and body (with the content of the article). In our paper we addressed keyword extraction in a cross-lingual setting: Koloski, Boshko, et al. "Out of Thin Air: Is Zero-Shot Cross-Lingual Keyword Detection Better Than Unsupervised?." arXiv preprint arXiv:2202.06650 (2022). [https://arxiv.org/pdf/2202.06650.pdf] For reproducing the results, you can use keyword datasets from the dataset http://hdl.handle.net/11356/1403 described in: Koloski, B., Pollak, S., Škrlj, B., & Martinc, M. (2021). Extending Neural Keyword Extraction with TF-IDF tagset matching. In: Proceedings of the EACL Hackashop on News Media Content Analysis and Automated Report Generation, Kiev, Ukraine, pages 22–29.

DataForSEO Labs API offers three powerful keyword research algorithms and historical keyword data:

• Related Keywords from the “searches related to” element of Google SERP. • Keyword Suggestions that match the specified seed keyword with additional words before, after, or within the seed key phrase. • Keyword Ideas that fall into the same category as specified seed keywords. • Historical Search Volume with current cost-per-click, and competition values.

Based on in-market categories of Google Ads, you can get keyword ideas from the relevant Categories For Domain and discover relevant Keywords For Categories. You can also obtain Top Google Searches with AdWords and Bing Ads metrics, product categories, and Google SERP data.

You will find well-rounded ways to scout the competitors:

• Domain Whois Overview with ranking and traffic info from organic and paid search. • Ranked Keywords that any domain or URL has positions for in SERP. • SERP Competitors and the rankings they hold for the keywords you specify. • Competitors Domain with a full overview of its rankings and traffic from organic and paid search. • Domain Intersection keywords for which both specified domains rank within the same SERPs. • Subdomains for the target domain you specify along with the ranking distribution across organic and paid search. • Relevant Pages of the specified domain with rankings and traffic data. • Domain Rank Overview with ranking and traffic data from organic and paid search. • Historical Rank Overview with historical data on rankings and traffic of the specified domain from organic and paid search. • Page Intersection keywords for which the specified pages rank within the same SERP.

All DataForSEO Labs API endpoints function in the Live mode. This means you will be provided with the results in response right after sending the necessary parameters with a POST request.

The limit is 2000 API calls per minute, however, you can contact our support team if your project requires higher rates.

We offer well-rounded API documentation, GUI for API usage control, comprehensive client libraries for different programming languages, free sandbox API testing, ad hoc integration, and deployment support.

We have a pay-as-you-go pricing model. You simply add funds to your account and use them to get data. The account balance doesn't expire.