To access the Text analysis dashboard of a dataset, click the pie chart icon at the end of the dataset element, or click the three dots next to it and choose ‘Dashboard’. Then, click on the Text analysis tab at the top. Once it opens, you can use the bar on the left side of the screen to navigate through its elements or add data series for comparison. You can also filter the data, drill down on the charts, change their display settings and export them.
a) Summary View
The goal of the summary view is to provide a general overview on your data, including the share of sentiment, the most frequently recognized positive and negative labels or Custom Labels.
The summary view contains the following charts:
- the distribution of mentions based on polarity,
- the distribution of Custom Labels shown in a pie chart (if there are Custom Labels applied to the dataset),
- the drivers and barriers identified in the verbatims: the most common co-mentions of entities and phrases, grouped by polarity (things that respondents really enjoyed and some of the main problems they had),
- top labels: the breakdown of automatically recognized labels and Custom Labels by frequency and opinion index - you can use the toggles at the top to hide Custom Labels and to display example mentions at the end of each row,
- a topic map, displaying the most common topics in a frame of reference, with mention numbers on the X-axis and Opinion Index on the Y-axis - you can use the toggles at the top to hide Custom Labels and to display example mentions above the topics,
- a label breakdown, which is a series of bar charts showing the most common topics and their most frequent co-mentions, plus the polarity of these co-mentions,
- a topic cloud of the most common topics (labels) or phrases (you can switch between these), where the size and color of the words reflects their frequency and polarity,
- a list of the top positive labels and another of the top negative labels, in both cases you can see their frequency, opinion index and the number of positive and negative mentions - you can also choose to hide Custom Labels or to add example mentions for each topic.
b) Relations View
The relations view helps you understand hidden connections in the dataset: looking at these charts, you’ll see the causalities and discover the correlations between the discovered topics.
It contains the following elements:
- a topic graph of the key topics of the dataset with their common connected positive and negative phrases – the numbers indicate their frequency,
- an associations table of the most significant co-mentions of labels and their polarity - you can choose to include example mentions,
- the key topics affecting the overall opinion: this is based on Pearson-correlation, the strength of which is also displayed next to the topics’ frequency and opinion index - you can choose to display example mentions,
- the top Custom Labels affecting the overall opinion (only if there are Custom Labels applied to the dataset),
- a key topics correlation map, a frame of reference in which the most common topics are placed based on the opinion index and the opinion correlation - you can opt to only display Custom Labels or automatically recognized labels.
c) Labels View
The labels view highlights the main entities and emotions recognized in your dataset.
It contains four tables, and in each case, the tables display the labels’ frequency, opinion index, and the number of positive and negative mentions.
- top brands,
- top emotions,
- top people.
d) Time Series View
The time series view organizes your data along a daily, weekly or monthly timeline, so that you can get an understanding of how trends might have shifted over different periods.
Its two charts place the distribution of opinions and the mention frequency of topics on a timeline.
However, if you have used Custom Labels for the dataset in question, the time series view will contain two time series charts for these as well: the Custom Label distribution and the Custom Label categories.
e) Categories View
The tables in the categories view break down predefined categories in your dataset according to their frequency, opinion index and the number of positive and negative mentions. The name of the tables contain the category that was used in each case. The number of charts depends on the number of categories in the dataset.
f) Custom Labels View
If you have applied Custom Labels in your dataset, this view will show you:
- their distribution on a pie chart,
- the distribution together with the Custom Label categories you have created, also a pie chart,
- a custom label map in a frame of reference where the X-axis shows mention numbers and the Y-axis shows the Opinion Index,
- a table of the top custom labels by categories, with the labels’ frequency, opinion index, and the number of positive and negative mentions.
g) Feed View & Drilldown
The feed view lists every single verbatim in your dataset: you can read them and see the recognized topics, as well as the positive and negative expressions found in them, highlighted in red and green. For each verbatim, an overall polarity score is displayed, and you can see the score for each label based on that particular verbatim as well. In the top right corner of each verbatim, you can find a cogwheel icon to access verbatim settings, and a language icon to translate the verbatim into English.
The feed view is searchable using the search bar at the top, and you can change the order of the verbatims, too: they can go from most positive to most negative, or vice versa. You can choose to export the verbatims using the export button at the top right corner of the view.
The feed view is also useful for editing custom labels and polarities of topics.
When you drill down on any topic or label anywhere on the Text Analysis Dashboard, you will also get a smaller feed view, listing all the verbatims under that label or topic.
h) NPS® View
This view will only show up in your dashboard if you have NPS® data in that particular dataset, to help you understand more about your Promoters, Passives and Detractors. Two charts will appear here.
The NPS® Topic Map displays topics in a frame of reference where the X-axis is the Net Promoter Score®, and the Y-axis is the Opinion Index.
- The NPS® Distribution For Top Labels is a bar chart, where top labels are broken down by Promoters, Passives and Detractors.
i) CSAT View
This view will only show up in your dashboard if you have CSAT data in that dataset. If you do, two charts appear in this view to help you understand more about the key reasons why your customers might be satisfied or dissatisfied.
The Customer Satisfaction Score Topic Map displays topics in a frame of reference where the X-axis is the CSAT Score, and the Y-axis is the Opinion Index.
The Customer Satisfaction Score Distribution For Top Labels is a bar chart, where top labels are broken down by satisfied, neutral and dissatisfied customers.
Read more about this feature on our blog:
Analyze your CSAT feedback with Zurvey.io and get more satisfied customers
j) Individual Marker View
This view only appears if your dataset contains Individual Marker data. By default, three different charts appear, however, if multiple Individual Marker columns are available in your dataset, charts are generated for each of them.
- The Individual Markers With The Most Positive Verbatims chart displays the Individual Markers that have the highest frequency of positive verbatims. Bar colors represent the average opinion index: darker green signifies a more positive sentiment, while lighter green suggests a less intense positivity. If you drill down in the chart, all verbatims that contributed to the average opinion index will be shown.
- The Individual Markers With The Most Negative Verbatims chart displays the Individual Markers that have the highest frequency of negative verbatims. Bar colors represent the average opinion index: darker red signifies a more negative sentiment, while lighter red indicates a milder negativity. If you drill down in the chart, all verbatims that contributed to the average opinion index will be shown.
- The Most Frequent Individual Markers chart reveals Individual Markers that have the highest frequency of verbatims. Each marker's colored sections indicate the sentiments — negative, neutral and positive.
