You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: docs/source/introduction/overview.rst
+67-11Lines changed: 67 additions & 11 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -6,13 +6,11 @@
6
6
Overview
7
7
========
8
8
9
-
ScrapeGraphAI is a open-source web scraping python library designed to usher in a new era of scraping tools.
10
-
In today's rapidly evolving and data-intensive digital landscape, this library stands out by integrating LLM and
11
-
direct graph logic to automate the creation of scraping pipelines for websites and various local documents, including XML,
12
-
HTML, JSON, and more.
9
+
ScrapeGraphAI is an **open-source** Python library designed to revolutionize **scraping** tools.
10
+
In today's data-intensive digital landscape, this library stands out by integrating **Large Language Models** (LLMs)
11
+
and modular **graph-based** pipelines to automate the scraping of data from various sources (e.g., websites, local files etc.).
13
12
14
-
Simply specify the information you need to extract, and ScrapeGraphAI handles the rest,
15
-
providing a more flexible and low-maintenance solution compared to traditional scraping tools.
13
+
Simply specify the information you need to extract, and ScrapeGraphAI handles the rest, providing a more **flexible** and **low-maintenance** solution compared to traditional scraping tools.
16
14
17
15
Why ScrapegraphAI?
18
16
==================
@@ -21,17 +19,75 @@ Traditional web scraping tools often rely on fixed patterns or manual configurat
21
19
ScrapegraphAI, leveraging the power of LLMs, adapts to changes in website structures, reducing the need for constant developer intervention.
22
20
This flexibility ensures that scrapers remain functional even when website layouts change.
23
21
24
-
We support many Large Language Models (LLMs) including GPT, Gemini, Groq, Azure, Hugging Face etc.
25
-
as well as local models which can run on your machine using Ollama.
22
+
We support many LLMs including **GPT, Gemini, Groq, Azure, Hugging Face** etc.
23
+
as well as local models which can run on your machine using **Ollama**.
26
24
27
25
Library Diagram
28
26
===============
29
27
30
-
With ScrapegraphAI you first construct a pipeline of steps you want to execute by combining nodes into a graph.
31
-
Executing the graph takes care of all the steps that are often part of scraping: fetching, parsing etc...
32
-
Finally the scraped and processed data gets fed to an LLM which generates a response.
28
+
With ScrapegraphAI you can use many already implemented scraping pipelines or create your own.
29
+
30
+
The diagram below illustrates the high-level architecture of ScrapeGraphAI:
ScrapeGraphAI is an open-source python library that uses large language models (LLMs) and graph logic to automate the creation of scraping pipelines for websites and various document types.
43
+
44
+
2. **How does ScrapeGraphAI differ from traditional scraping tools?**
45
+
46
+
Traditional scraping tools rely on fixed patterns and manual configurations, whereas ScrapeGraphAI adapts to website structure changes using LLMs, reducing the need for constant developer intervention.
47
+
48
+
3. **Which LLMs are supported by ScrapeGraphAI?**
49
+
50
+
ScrapeGraphAI supports several LLMs, including GPT, Gemini, Groq, Azure, Hugging Face, and local models that can run on your machine using Ollama.
51
+
52
+
4. **Can ScrapeGraphAI handle different document formats?**
53
+
54
+
Yes, ScrapeGraphAI can scrape information from various document formats such as XML, HTML, JSON, and more.
55
+
56
+
5. **I get an empty or incorrect output when scraping a website. What should I do?**
57
+
58
+
There are several reasons behind this issue, but for most cases, you can try the following:
59
+
60
+
- Set the `headless` parameter to `False` in the graph_config. Some javascript-heavy websites might require it.
61
+
62
+
- Check your internet connection. Low speed or unstable connection can cause the HTML to not load properly.
63
+
64
+
- Try using a proxy server to mask your IP address. Check out the :ref:`Proxy` section for more information on how to configure proxy settings.
65
+
66
+
- Use a different LLM model. Some models might perform better on certain websites than others.
67
+
68
+
- Set the `verbose` parameter to `True` in the graph_config to see more detailed logs.
69
+
70
+
- Visualize the pipeline graphically using :ref:`Burr`.
71
+
72
+
If the issue persists, please report it on the GitHub repository.
73
+
74
+
6. **How does ScrapeGraphAI handle the context window limit of LLMs?**
75
+
76
+
By splitting big websites/documents into chunks with overlaps and applying compression techniques to reduce the number of tokens. If multiple chunks are present, we will have multiple answers to the user prompt, and therefore, we merge them together in the last step of the scraping pipeline.
77
+
78
+
7. **How can I contribute to ScrapeGraphAI?**
79
+
80
+
You can contribute to ScrapeGraphAI by submitting bug reports, feature requests, or pull requests on the GitHub repository. Join our `Discord <https://discord.gg/uJN7TYcpNa>`_ community and follow us on social media!
0 commit comments