delab-trees

Delab Trees

A library to analyze conversation trees.

Installation

pip install delab_trees

Get started

Example data for Reddit and Twitter are available here https://github.com/juliandehne/delab-trees/raw/main/delab_trees/data/dataset_[reddit|twitter]_no_text.pkl. The data is structure only. Ids, text, links, or other information that would break confidentiality of the academic access have been omitted.

The trees are loaded from tables like this:

	tree_id	post_id	parent_id	author_id	text	created_at
0	1	1	nan	james	I am James	2017-01-01 01:00:00
1	1	2	1	mark	I am Mark	2017-01-01 02:00:00
2	1	3	2	steven	I am Steven	2017-01-01 03:00:00
3	1	4	1	john	I am John	2017-01-01 04:00:00
4	2	1	nan	james	I am James	2017-01-01 01:00:00
5	2	2	1	mark	I am Mark	2017-01-01 02:00:00
6	2	3	2	steven	I am Steven	2017-01-01 03:00:00
7	2	4	3	john	I am John	2017-01-01 04:00:00

This dataset contains two conversational trees with four posts each.

Currently, you need to import conversational tables as a pandas dataframe like this:

import pandas as pd
from delab_trees import TreeManager

d = {'tree_id': [1] * 4,
     'post_id': [1, 2, 3, 4],
     'parent_id': [None, 1, 2, 1],
     'author_id': ["james", "mark", "steven", "john"],
     'text': ["I am James", "I am Mark", " I am Steven", "I am John"],
     "created_at": [pd.Timestamp('2017-01-01T01'),
                    pd.Timestamp('2017-01-01T02'),
                    pd.Timestamp('2017-01-01T03'),
                    pd.Timestamp('2017-01-01T04')]}
df = pd.DataFrame(data=d)
manager = TreeManager(df) 
# creates one tree
test_tree = manager.random()

Note that the tree structure is based on the parent_id matching another rows post_id.

You can now analyze the reply trees basic metrics:

from delab_trees.main import get_test_tree
from delab_trees.delab_tree import DelabTree

test_tree : DelabTree = get_test_tree()
assert test_tree.total_number_of_posts() == 4
assert test_tree.average_branching_factor() > 0

A summary of basic metrics can be attained by calling

from delab_trees.main import get_test_tree
from delab_trees.delab_tree import DelabTree

test_tree : DelabTree = get_test_tree()
print(test_tree.get_author_metrics())

# >>> removed [] and changed {} (merging subsequent posts of the same author)
# >>>{'james': <delab_trees.delab_author_metric.AuthorMetric object at 0x7fa9c5496110>, 'steven': <delab_trees.delab_author_metric.AuthorMetric object at 0x7fa9c5497dc0>, 'john': <delab_trees.delab_author_metric.AuthorMetric object at 0x7fa9c5497a00>, 'mark': <delab_trees.delab_author_metric.AuthorMetric object at 0x7fa9c5497bb0>}

More complex metrics that use the full dataset for training can be gotten by the manager:

import pandas as pd
from delab_trees import TreeManager

d = {'tree_id': [1] * 4,
     'post_id': [1, 2, 3, 4],
     'parent_id': [None, 1, 2, 1],
     'author_id': ["james", "mark", "steven", "john"],
     'text': ["I am James", "I am Mark", " I am Steven", "I am John"],
     "created_at": [pd.Timestamp('2017-01-01T01'),
                    pd.Timestamp('2017-01-01T02'),
                    pd.Timestamp('2017-01-01T03'),
                    pd.Timestamp('2017-01-01T04')]}
df = pd.DataFrame(data=d)
manager = TreeManager(df) # creates one tree
rb_vision_dictionary : dict["tree_id", dict["author_id", "vision_metric"]] = manager.get_rb_vision()

The following two complex metrics are implemented:

from delab_trees.main import get_test_manager

manager = get_test_manager()
rb_vision_dictionary = manager.get_rb_vision() # predict an author having seen a post
pb_vision_dictionary = manager.get_pb_vision() # predict an author to write the next post

How to cite

    @article{dehne_dtrees_23,
    author    = {Dehne, Julian},
    title     = {Delab-Trees: measuring deliberation in online conversations},        
    url = {https://github.com/juliandehne/delab-trees}     
    year      = {2023},
}