CONVERSATION HIGHLIGHTS

By analyzing the hashtags:

"#staysafe","#domesticviolence","#DomesticAbuse", "#whyistayed", "#whyileft",#metoo, "#HEFORSHEATHOME","#womencount","antidomesticviolenceduringepidemic", "Mask-19"

The trend has elevated since the lockdown has started:

March 12,2020

After realizing the pick of conversation rose on March 12th, when lockdown started in some countries, the research focus on this data, analyzing 105k mentions:

What is the sentiment and most common words?

55%

28%

17%

Analysis of hashtags: "#staysafe","#domesticviolence"& "#DomesticAbuse"

Positive Sentiment:

Negative Sentiment:

Women that tweet about these hashtags and are marked as positive, using "Vader Sentiment Library" in Python, the main topic is women seeking support about their past experiences, another highlight of the words "someone known" in both positive and negative wordings.

Analysis of hashtags: "#WHYILEFT","#WHYISTAYED"

Positive Sentiment:

Negative Sentiment:

With the analyzed hashtags, which were part of an initiative that has been present since 2014, stand out movements that have been related to the topic:

#metoo,#heforshe,#yesallwomen #timesup,#iwillgoout,#bringackourgirls, and the most recently #niunamenos, lead by México in March, also the organization "orange the world" stands out which is a campaign from UNESCO fighting violence against women.

Discover more

Our Mission

Due to the ambiguity of Sentiment Analysis, especially in this subject where positive is not positive, in the subject where almost every keyword is negative, we did a further Topic Modeling Analysis, thus understanding not only the keywords but the probabilistic relationship between them.

Relevant topics of conversation

Through an analysis of NLP, using Topic Modeling, which is a text mining technique that provides methods for identifying co-occurring keywords.

It helps in discovering hidden topics in the document, annotate the documents with these topics, and organize a large amount of unstructured data.

We used the Latent Dirichlet Allocation probabilistic model, it worked by mapping each document in our corpus to a set of topics that covers a good deal of the words in the document. What LDA does in order to map the documents to a list of topics is assign topics to arrangements of words, e.g. n-grams, by building a topic per document model and words per topic model, modeled as Dirichlet distributions.

Main Topics :

Topic 1: Is related to topic three, because both are from the #WhyIleft movement, this topic is related to the International Women´s Day and the fight towards an abusive relationship and harassment.
Topic 2: Metoo movement supporting the survivors' victims, quoting "kid" and how "leave" as well "fight" and "evidence", concluding that these factors are important for these people telling their stories.
Topic 3: #WhyIleft, but in the context of sharing supports towards the victim´s and stopping violence, creating a community and promote awareness
Topic 4: #Metoomovement: Advocating the movement and uniting with the #believeinwomen,#timesup, it is also about "tara_reade", which is an American News that went out this week where she blames Joe Biden of domestic abuse.
Topic 5: Campaign of UN Women, towards ending violence and rape.
Topic 6: Providing help to various types of abuse: psychological, punished innocent and wanting to design a solution is also the only topic that has several bigrams (6)

Twitter Data Takeaway:

The nature of the social network, the main authors are organizations, influencers or leaders of opinion, where most of the conversation is of RT, but we can identify clear topics that would have been impossible with sentiment analysis, in which these organizations invite people to share their story with a certain hashtag, and we can identify the topics revolving the movements and how it changes reflecting on the most important subjects that people write.

Even though people do share their story, there are no testimonies, that is the reason we also analyzed Reddit, a good recommendation will be for organizations to pursue by this hashtags to tell their story in Reddit, thus understanding their needs better.

CONVERSATION HIGHLIGHTS

March 12,2020

What is the sentiment and most common words?

Relevant topics of conversation

Through an analysis of NLP, using Topic Modeling, which is a text mining technique that provides methods for identifying co-occurring keywords.

It helps in discovering hidden topics in the document, annotate the documents with these topics, and organize a large amount of unstructured data.

Main Topics :

Topic 1: Is related to topic three, because both are from the #WhyIleft movement, this topic is related to the International Women´s Day and the fight towards an abusive relationship and harassment.

Topic 2: Metoo movement supporting the survivors' victims, quoting "kid" and how "leave" as well "fight" and "evidence", concluding that these factors are important for these people telling their stories.

Topic 3: #WhyIleft, but in the context of sharing supports towards the victim´s and stopping violence, creating a community and promote awareness

Topic 4: #Metoomovement: Advocating the movement and uniting with the #believeinwomen,#timesup, it is also about "tara_reade", which is an American News that went out this week where she blames Joe Biden of domestic abuse.

Topic 5: Campaign of UN Women, towards ending violence and rape.

Topic 6: Providing help to various types of abuse: psychological, punished innocent and wanting to design a solution is also the only topic that has several bigrams (6)

Twitter Data Takeaway:

​

Even though people do share their story, there are no testimonies, that is the reason we also analyzed Reddit, a good recommendation will be for organizations to pursue by this hashtags to tell their story in Reddit, thus understanding their needs better.