First Published: February 6, 2025
Our data is currently stored in a data file. That data was gathered by fetching all talks from President Nelson from 1984 to the present. They were then put through a natural language processing algorithm to identify key words and track word usage.
with open('talk_data.json', 'r') as f:
talk_data = json.load(f)
chart_data = prepare_chart_data(talk_data)
df = create_dataframe_from_chart_data(chart_data)
df.head()| word | date | count | |
|---|---|---|---|
| 0 | President | 2025-04 | 1 |
| 1 | President | 2024-10 | 1 |
| 2 | President | 2024-04 | 3 |
| 3 | President | 2023-10 | 2 |
| 4 | President | 2023-04 | 3 |
Due to the way the text is broken out into tokens, words that are commonly used together like ‘Jesus’ and ‘Christ’ need to be combined to ensure they are not counted twice. Additionally the words ‘Lord’ and ‘Savior’ are added combined with the final term ‘Jesus Christ’. This was done on a select few other pairs.
average_word_mapping = {
'jesus christ': ['Jesus', 'Christ'],
}
combined_df = average_words(df, average_word_mapping)
combine_word_mapping = {
'Jesus christ': ['jesus christ', 'Lord'],
'Jesus Christ': ['Jesus christ', 'Savior'],
'God': ['God', 'Father'],
'Eternal Life': ['life', 'eternal'],
'Temple': ['temple', 'house']
}
combined_df = combine_words(combined_df, combine_word_mapping)In order to identify the most commonly used words, we will apply a group-by clause on the data-frame. This combines the counts of all the words from all the different years, and return then top 10 in descending order.
top_n = 10
top_words_df = top_n_words(combined_df, top_n)
top_words_df.head(top_n)word
Jesus Christ 1345.0
God 901.0
Eternal Life 765.0
Temple 458.0
day 446.0
Church 421.0
child 378.0
family 376.0
time 374.0
love 373.0
Name: count, dtype: float64The top 10 most used words by President Russell M. Nelson in descending order from 1984 to 2024.
Now we can plot or clean and sorted data to see word usage over time.
The following chart has data filtered from 2018 onward, even though the actual data goes back until 1984. This is done becuase Russel M. Nelson was sustained as President of The Church of Jesus Christ of Latter-day Saints in January of 2018.
We are able to see that while there are fluctuations in the usage of ‘Jesus Christ’ over time, usage of it is consistently high. The most noticeable spike in October 2024 when President Nelson gave his most recent General Conference address The Lord Jesus Christ Will Come Again.
Data collected, and report completed by Ethan Glenn. View on GitHub