Análisis de tweets atacando a @guadavazquez

Author

Fernanda Aguirre

Published

January 16, 2024

Datos

Code
df_attacks = attacks.loc[attacks["to_journalist"].isin([USERNAME])]
print(f"Número de ataques: {len(df_attacks)}")
Número de ataques: 216

Frecuencia de ataques

Proporción de ataques = (Número de ataques / Número de menciones) * 100

Code
journalist_mentions = len(df.loc[df["to_journalist"].isin([USERNAME])])
journalist_attacks = len(df_attacks)

percentage_attacks = (journalist_attacks / journalist_mentions) * 100

proportion = (percentage_attacks / 100) * 10
proportion_rounded = round(proportion, 1)


print(
    f"Aproximadamente {proportion_rounded} de cada 10 publicaciones que mencionan a {USERNAME} son ataques"
)
Aproximadamente 0.6 de cada 10 publicaciones que mencionan a @guadavazquez son ataques

Frecuencia de los ataques en función del número de seguidores

Proporción de ataques por seguidor = Número de ataques / Número de seguidores

Code
proportion_followers = (len(df_attacks) / 230000) * 1000
formatted_proportion = "{:.2f}".format(proportion_followers)
print(f"Por cada 1K seguidores, aproximadamente hubo {formatted_proportion} ataques para {USERNAME}")
Por cada 1K seguidores, aproximadamente hubo 0.94 ataques para @guadavazquez

Ranking de tipos de ataques

Code
conditions = [
    "women",
    "politics",
    "appearance",
    "racism",
    "class",
    "lgbti",
    "criminal",
    "calls",
]
attacks_count = df_attacks[conditions].sum()
attacks_count = attacks_count.sort_values(ascending=False)
attacks_count
women         91
politics      46
appearance    23
racism        13
calls          7
class          6
lgbti          6
criminal       2
dtype: int64

Número de ataques por tipo de evento

Code
df_attacks["event"].value_counts()
event
1er debate              63
elecciones balotaje     53
elecciones generales    42
2do debate              38
debate balotaje         20
Name: count, dtype: int64

Publicaciones por evento

Code
journalist_posts = df.loc[df["from_journalist"].isin([USERNAME])]
journalist_posts = journalist_posts.dropna(subset=["from_journalist"])

eventos = [
    "1er debate",
    "2do debate",
    "elecciones generales",
    "debate balotaje",
    "elecciones balotaje",
]
colors = ["green", "purple", "orange", "red", "blue"]
eventos_count = {}

fig = px.line()

for i, evento in enumerate(eventos):
    evento_data = journalist_posts.loc[journalist_posts["event"].isin([evento])]
    evento_count = evento_data.groupby("dt_date").size().reset_index(name="count")
    eventos_count[evento] = evento_count
    fig.add_scatter(
        x=evento_count["dt_date"],
        y=evento_count["count"],
        name=evento,
        line=dict(color=colors[i]),
        hovertemplate="posts: %{y}",
    )

fig.update_layout(title=f"Publicaciones de {USERNAME}", width=1000)
fig.update_xaxes(type="category")
fig.update_yaxes(range=[0, 100])
fig.show()

Hashtags

Code
df_attacks["hashtags"] = df_attacks["text"].apply(
    lambda x: (
        np.nan
        if pd.isnull(x) or not isinstance(x, str) or len(re.findall(r"#\w+", x)) == 0
        else re.findall(r"#\w+", x)
    )
)

df_attacks["hashtags"] = df_attacks["hashtags"].apply(
    lambda x: ", ".join(x) if isinstance(x, list) else x
)

# convert dataframe column to list
hashtags = df_attacks["hashtags"].unique()

# remove nan items from list
hashtags = [x for x in hashtags if not pd.isna(x)]

# split items into a list based on a delimiter
hashtags = [x.split(",") for x in hashtags]

# flatten list of lists
hashtags = [item for sublist in hashtags for item in sublist]

# remove whitespaces
hashtags = list(map(lambda x: x.replace(" ", ""), hashtags))

# count items on list
hashtags_count = pd.Series(hashtags).value_counts()

hashtags_count
#Tenemos    1
Name: count, dtype: int64

Menciones

Code
df_attacks["mentions"] = df_attacks["text"].apply(
    lambda x: (
        np.nan
        if pd.isnull(x) or not isinstance(x, str) or len(re.findall(r"@(\w+)", x)) == 0
        else re.findall(r"@(\w+)", x)
    )
)

df_attacks["mentions"] = df_attacks["mentions"].apply(
    lambda x: ", ".join(x) if isinstance(x, list) else x
)

# convert dataframe column to list
mentions = df_attacks["mentions"].unique()

# remove nan items from list
mentions = [x for x in mentions if not pd.isna(x)]

# split items into a list based on a delimiter
mentions = [x.split(",") for x in mentions]

# flatten list of lists
mentions = [item for sublist in mentions for item in sublist]

# remove whitespaces
mentions = list(map(lambda x: x.replace(" ", ""), mentions))

# count items on list
mentions_count = pd.Series(mentions).value_counts()

mentions_count
vivicanosaok      1
guadavazquez      1
edufeiok          1
EsmeraldaMitre    1
Name: count, dtype: int64

Tokens

Code
# load the spacy model for Spanish
nlp = spacy.load("es_core_news_sm")

# load stop words for Spanish
STOP_WORDS = nlp.Defaults.stop_words


# Function to filter stop words
def filter_stopwords(text):
    # lower text
    doc = nlp(text.lower())
    # filter tokens
    tokens = [
        token.text
        for token in doc
        if not token.is_stop and token.text not in STOP_WORDS and token.is_alpha
    ]
    return " ".join(tokens)


# apply function to dataframe column
df_attacks["text_pre"] = df_attacks["text"].apply(filter_stopwords)

# count items on column
token_counts = df_attacks["text_pre"].str.split(expand=True).stack().value_counts()[:20]

token_counts
gato         25
guada        17
q            14
jajaja       13
vos          10
zurdos        8
sos           8
cara          7
milei         7
medicado      6
abrazo        6
peronchos     6
entiendo      5
gatos         5
re            5
familia       5
acá           5
bregman       5
zurda         5
asco          5
Name: count, dtype: int64