Phd - SLR - Sistemātiskais Literatūras apskats

 

 

Examples of SLR tables:

  1. https://puupuls.notion.site/SLR-Tabula-1-04cafc5ed342497b92629b87c2c1e548

    1. Missing:

      1. Cite by year

      2. Future research

  2. https://www.evernote.com/shard/s736/sh/2c598470-3b41-6a1c-d629-3fa44ec1e0c3/c340ef470d5f0d9391b709d03751445f

 

Examples of Finished SLR:

  1. http://share.yellowrobot.xyz/quick/2023-2-28-62301A6A-9494-43EE-B86B-007AF6F685EB.pdf

  2. https://sci-hub.hkvisa.net/10.1016/j.infsof.2008.09.009

  3. https://www.sciencedirect.com/science/article/pii/S0950584908001390

  4. http://share.yellowrobot.xyz/quick/2023-2-28-0CCAE1B2-6579-4F45-97E3-A12AEDA732B0.pdf

 

References shpuld be taken from (Do not use arxiv as reference)

  1. https://aclanthology.org

  2. https://www.semanticscholar.org

 

Create and share with me SLR (Systematic Literature Review) table using Notion, Google Sheets etc:

  1. Find datastets that have not yet reached 99-95% accuracy

  2. Find SOTA (State of the Art) metrics and scalar values for each dataset

  3. Find active research topics - data pre-processing, result post-processing, loss functions, data augementations, model architectures

  4. Table columns

    1. No. At least 30 scientific publictions

    2. Year

    3. Title

    4. Keywords

    5. Link to publication

    6. Link to Github if available

    7. Otganization (university, company)

    8. Citation count (Semantic schollar)

    9. Data set (which have been used) - If there is no published information or data set information: NONE

    10. Metrics and Results (used to determine the method performance on each dataset), if there are no exact results, then: NONE

    11. Model (what is the model architecture), can also be a NON-DL model

    12. Method (what is the main method being used)

    13. Further Research (what is recommended for further study)

 

Best tool for reading PDFs (or find others using langchain):

  1. https://chatpdf.com

  2. perplexity.ai PDF upload function

 

 

Tool to access any publication:

  1. http://sci-hub.hkvisa.net

 

Use these databases for searching papers:

  1. http://semanticscholar.org

  2. http://researchrabbitapp.com

  3. https://deeplearn.org

  4. https://scirate.com

  5. https://arxiv-sanity-lite.com

  6. http://perplexity.ai (academic section)

  7. https://typeset.io

 

Obligāti datu kopām pārmeklēt līdzīgas ar:

  1. http://zenodo.org

  2. http://paperswithcode.com

 

Rezultātus starp modeļiem salīdzināt apmēram šādi: image-20240105115752204