amandakonet commited on
Commit
fedacac
β€’
1 Parent(s): 3404fa4

start entailment

Browse files
Files changed (1) hide show
  1. app.py +16 -6
app.py CHANGED
@@ -9,7 +9,7 @@ st.title('Combatting Climate Change Misinformation with Transformers')
9
 
10
  # section 1: the context, problem; how to address
11
  st.markdown("### The Problem πŸ€”")
12
- st.markdown("")
13
 
14
 
15
  # section 2: what is misinformation? how is it combatted now? how successful is this?
@@ -38,7 +38,7 @@ st.markdown("### How can Transformers Help?")
38
 
39
  # fever dataset
40
  # entailment/contradiction/neutral adoption to support/refute/n.e.i.
41
- # extention to climate
42
 
43
  # section 4: The process
44
  # this is the pipeline in my notes (u are here highlight)
@@ -50,10 +50,9 @@ st.markdown("1. User inputs a climate claim")
50
  #input_gif = Image.open('images/input_box.gif')
51
  #st.image(input_gif, width=100)
52
 
53
- st.markdown("2. Retrieve evidence related to input claim")
54
- st.markdown(" - Similarity between claim and available documents")
55
- st.markdown(" - For each claim, collect N related documents")
56
- st.markdown(" - Current area of research: How do we keep the set of curated documents up-to-date? Validate their contents?")
57
 
58
  st.markdown("3. Send (claim, evidence) pairs to a transformer model. Have the model predict whether each evidence supports, refutes, or is not relevant to the claim. (πŸ“ YOU ARE HERE!)")
59
 
@@ -61,6 +60,17 @@ st.markdown("4. Report back to the user: The supporting evidence for the claim (
61
 
62
 
63
  # section 5: my work
 
 
 
 
 
 
 
 
 
 
 
64
 
65
  # section 6: analysis
66
 
 
9
 
10
  # section 1: the context, problem; how to address
11
  st.markdown("### The Problem πŸ€”")
12
+ st.markdown("Misinformation about climate change spreads quickly and has direct impacts on public opinion and public policy surrounding the climate. Further, misinformation is difficult to combat, and people are able to \"verify\" false climate claims on biased sites. Ideally, people would be able to easily verify climate claims. This is where transformers come in.")
13
 
14
 
15
  # section 2: what is misinformation? how is it combatted now? how successful is this?
 
38
 
39
  # fever dataset
40
  # entailment/contradiction/neutral adoption to support/refute/n.e.i.
41
+ # extention to climate -climatefever
42
 
43
  # section 4: The process
44
  # this is the pipeline in my notes (u are here highlight)
 
50
  #input_gif = Image.open('images/input_box.gif')
51
  #st.image(input_gif, width=100)
52
 
53
+ st.markdown("2. Retrieve evidence related to input claim \
54
+ - For each claim, collect N related documents. These documents are selected by finding the N documents with the highest similarity scores to the claim.")
55
+ st.markdown("- Current area of research: How do we keep the set of curated documents up-to-date? Validate their contents?")
 
56
 
57
  st.markdown("3. Send (claim, evidence) pairs to a transformer model. Have the model predict whether each evidence supports, refutes, or is not relevant to the claim. (πŸ“ YOU ARE HERE!)")
58
 
 
60
 
61
 
62
  # section 5: my work
63
+ st.markdown("### Climate Claim Fact-Checking with Transformers")
64
+
65
+ st.markdown("My work focuses on step 3 of the process: Training a transformer model to accurately categorize (claim, evidence) as:")
66
+ st.markdown("* evidence *supports* (entails) claim")
67
+ st.markdown("* evidence *refutes* (contradicts) claim")
68
+ st.markdown("* evidence *does not provide enough info to support or refute* (neutral) claim")
69
+ st.markdown("For this project, I fine-tuned 3 different models on the text entailment task.")
70
+
71
+ st.markdown("")
72
+
73
+
74
 
75
  # section 6: analysis
76