Lesson Report:
## 1) Title
**Launching the Social Media Verification Unit: Image Geolocation & Reverse Image Search (Hands-on Exercises)**
The session introduced the next two-week course unit focused on analyzing social media content without “polluting� students’ personal recommendation algorithms. The instructor framed the unit around real-world misinformation problems (recycled/misattributed images) and began practical skills training in **image verification**, emphasizing **geolocation** and **reverse image search**, followed by two timed photo-location exercises.
—
## 2) Attendance
– **Students explicitly mentioned absent:** 0
– **Names of students who submitted answers / were mentioned as participating (for instructor reference):** Imat, Ruslan, Helen, Subhan, Natalia, Samira, Chinara, Yvonne, Kaneke, Floron, Ofarid, Nazbikia, Kamila, Sven, Danjek, Zamira (plus “Akaliâ€� asked for screen share/help accessing image).
—
## 3) Topics Covered (Chronological, with activity progression)
### A. Unit Preview & Setup: Social Media Module (Next Two Weeks)
– **Announcement of upcoming focus:** The class is transitioning into a two-week deep dive on **social media**, with significant time spent **searching through hashtags**.
– **Practical preparation request (important constraint):**
– Students were asked—if feasible and with no objections—to create **fresh/new accounts** on **Instagram and/or Twitter/X**.
– **Rationale:** Using personal accounts for repeated hashtag searching and content exploration can permanently distort students’ platform algorithms (“destroy your algorithms foreverâ€�).
– **Contingency plan:** Students who cannot make new accounts were asked to **notify the instructor** so accommodations can be arranged.
### B. Framing the Problem: Why Verification Matters (Theory → Real World)
– The instructor contextualized the verification skillset by describing a recurring misinformation pattern:
– **Actors (including states)** have repeatedly used **visual materials** (photos) that **do not align** with the claims attached to them.
– Example pattern: A photo shared as evidence from a current war zone is later found to be:
– taken **somewhere else**, and/or
– taken **years earlier**.
– These “recycled photosâ€� can be amplified by algorithms and accepted as authentic by everyday users.
– Connection to course goals:
– The second part of the course will rely heavily on students’ **independent ability to assess veracity** (how true/credible a claim is).
### C. Core Skill Introduction: Image Verification (Two Main Techniques)
The instructor introduced two primary techniques students will use to verify images encountered online:
1. **Geolocation**
– Definition (as taught): Linking an image to **as close to exact coordinates as possible**—the goal is to identify where the photo was taken with maximum precision.
2. **Reverse Image Search**
– Students were told this tool would be used for **multiple tasks** during the session and beyond.
– Reverse image search was treated as a foundational workflow for tracing image reuse and origin.
### D. Activity 1: “Cold Open� Photo Geolocation Challenge (Individual, Timed)
– **Objective:** Determine where the instructor’s photo was taken (as precisely as possible).
– **Format decision:** Initially considered groups of three, but switched to **no groups** for speed.
– **Timing:** ~5 minutes of independent work.
– **Scoring rubric (progressively more specific location = more points):**
– **Country** = 1 point
– **State/region** within country = 2 points
– **Town** = 3 points
– (The instructor later repeated an expanded rubric inconsistently, including “cityâ€� multiple times; the main functioning rubric used in grading was country/state/town.)
– **Logistics & troubleshooting:**
– Image was posted in the Zoom chat; some students could not access it.
– Instructor responded by:
– uploading to an image-sharing site and sending a **link**,
– then **screen-sharing** to ensure students could view the photo.
– Students were asked not to post answers immediately to avoid influencing others; answers were collected after the work period.
#### Activity 1 Results & Debrief
– **Correct location revealed by instructor:**
– **Country:** United States
– **State/region:** New York
– **Town:** Guilderland or Schenectady (both accepted as correct)
– **Student methods discussed (explicitly surfaced as acceptable workflows):**
– **Imat:** Downloaded image → used **Yandex Images** → compared returned options; also used **ChatGPT** (after finding a street name) to help narrow down location.
– Instructor emphasized Yandex Images is “super powerfulâ€� for this purpose and that using it is **not cheating**—it is a legitimate verification tool.
– **Helen:** Used **Google reverse image search** to identify U.S. and New York; then used **ChatGPT** to refine/confirm the area.
– **Natalia:** Commented that **ChatGPT gave a clear answer** (noted as part of tool discussion).
– **Key learning point from instructor: what made this image “easyâ€�**
– The photo contained **text/words**, specifically a **road/street sign**.
– Street names are often one of the fastest anchors for geolocation, especially when clearly visible and centered.
### E. Activity 2: Second Photo Geolocation Challenge (More Difficult)
– **Objective:** Repeat the location-identification task with a more challenging image.
– **Scoring rubric reiterated:**
– **Country** = 1 point
– **State/region** = 2 points
– **Town** = 3 points
– **Instructor note:** For students “in the know,â€� there is a “certain taskâ€� that can determine the location very precisely (not fully explained before transcript ends—likely intended as a lead-in to a specific method/tool).
– **Access/logistics issues and support:**
– A student asked for screen share; instructor noted screen sharing wouldn’t provide the needed advantage and emphasized students should be able to **download** the image to work with it.
– Link was shared; at least one student requested it to be sent again after switching devices (laptop).
– Confirmation that the link worked for some students.
– **Transcript ends before:**
– answers were collected,
– correct location was revealed,
– methods were debriefed.
—
## 4) Actionable Items (Outside-class follow-ups, organized by urgency)
### Urgent / Before Next Class (Preparation for the Social Media Unit)
– **Students:** Create a **fresh Instagram and/or Twitter(X)** account(s) if possible to avoid algorithm contamination during hashtag-based exercises.
– **Students who cannot create new accounts:** **Notify instructor** so alternatives/accommodations can be arranged.
### High Priority (Instructional Follow-up Needed)
– **Complete Activity 2 loop next session:**
– Collect/confirm student answers (if not already done off-transcript),
– Reveal the correct location,
– Debrief the “certain taskâ€�/technique hinted at for precise geolocation.
– **Accessibility/logistics:** Provide a stable, consistently accessible method for distributing images (e.g., shared folder/LMS post) since multiple students had chat/link viewing issues.
### Medium Priority (Course Skill Development)
– Plan a short mini-lesson comparing verification tools surfaced by students:
– **Yandex Images vs. Google reverse image search** (strengths/weaknesses, when to use each),
– how/when **AI tools (e.g., ChatGPT)** can assist (and limitations/risk of over-trusting AI outputs).
– Reinforce best practice: document verification steps (what clue was used—e.g., street sign text—and what tool confirmed it) to support reproducibility and credibility.
### Low Priority / Notes for Future Lessons
– Consider clarifying/standardizing the geolocation “pointâ€� rubric (country/state/town/city) to avoid confusion (rubric was repeated with inconsistent “cityâ€� tiers during instructions).
Homework Instructions:
ASSIGNMENT #1: Create a fresh social-media account for the social media/hashtag verification unit
You will set up a new (separate) account on Instagram and/or Twitter/X so you can participate in the next two weeks of hashtag searching and image-verification activities without “destroying your algorithms� on your personal account(s), as discussed in class (“I would ask you to make a new account on either or both Instagram and or Twitter… [otherwise] it will destroy your algorithms forever.�).
Instructions:
1. Decide which platform(s) you will use:
1) Create a new Instagram account, and/or
2) Create a new Twitter/X account (the professor noted “Twitter, X, whatever you want to call it�).
2. If you are unable or unwilling to create a new account, contact the professor as soon as possible:
– This was explicitly requested in class: “If you can’t do that, then please let me know and we’ll do our best to make things work.â€�
3. Create a *new* account (do not use your existing personal account) on the platform(s) you chose:
– Use the platform’s standard sign-up process (mobile app or web).
– Choose a username that does not reveal personal/sensitive information unless you want it to.
4. Keep the account “clean� for research purposes (so hashtag searching results are less biased by your prior behavior):
– Do not follow a lot of personal-interest accounts.
– Avoid liking/saving/reposting content unrelated to class activities.
– If prompted to follow suggested accounts during setup, skip or minimize this step when possible.
5. Adjust basic privacy and security settings in a way that supports class use:
– Set a strong password and store it securely.
– Enable two-factor authentication (2FA) if available.
– Decide whether you want the account to be public or private; either is fine unless your professor later specifies otherwise.
6. Make sure you can access the account reliably for in-class work:
– Confirm you can log in on the device you typically use for class.
– Save your login information somewhere you can access during class.
7. Bring the account(s) to the next class session(s) during this unit:
– You will be using these accounts for “quite a bit of searching through hashtags for the next two weeksâ€� and for skills tied to verification (including image verification using reverse image search and geolocation).